Title
string
Abstract
string
Status
string
User
string
text
string
label
int64
combined_text
string
__index_level_0__
int64
Adversarial Robustness of Deep Learning: Theory, Algorithms, and Applications
This tutorial aims to introduce the fundamentals of adversarial robustness of deep learning, presenting a well-structured review of up-to-date techniques to assess the vulnerability of various types of deep learning models to adversarial examples. This tutorial will particularly highlight state-of-the-art techniques in adversarial attacks and robustness verification of deep neural networks (DNNs). We will also introduce some effective countermeasures to improve the robustness of deep learning models, with a particular focus on adversarial training. We aim to provide a comprehensive overall picture about this emerging direction and enable the community to be aware of the urgency and importance of designing robust deep learning models in safety-critical data analytical applications, ultimately enabling the end-users to trust deep learning classifiers. We will also summarize potential research directions concerning the adversarial robustness of deep learning, and its potential benefits to enable accountable and trustworthy deep learning-based data analytical systems and applications.
Liked
zrz@andrew.cmu.edu
Adversarial Robustness of Deep Learning: Theory, Algorithms, and Applications : This tutorial aims to introduce the fundamentals of adversarial robustness of deep learning, presenting a well-structured review of up-to-date techniques to assess the vulnerability of various types of deep learning models to adversarial examples. This tutorial will particularly highlight state-of-the-art techniques in adversarial attacks and robustness verification of deep neural networks (DNNs). We will also introduce some effective countermeasures to improve the robustness of deep learning models, with a particular focus on adversarial training. We aim to provide a comprehensive overall picture about this emerging direction and enable the community to be aware of the urgency and importance of designing robust deep learning models in safety-critical data analytical applications, ultimately enabling the end-users to trust deep learning classifiers. We will also summarize potential research directions concerning the adversarial robustness of deep learning, and its potential benefits to enable accountable and trustworthy deep learning-based data analytical systems and applications.
1
zrz@andrew.cmu.edu [SEP] Adversarial Robustness of Deep Learning: Theory, Algorithms, and Applications : This tutorial aims to introduce the fundamentals of adversarial robustness of deep learning, presenting a well-structured review of up-to-date techniques to assess the vulnerability of various types of deep learning models to adversarial examples. This tutorial will particularly highlight state-of-the-art techniques in adversarial attacks and robustness verification of deep neural networks (DNNs). We will also introduce some effective countermeasures to improve the robustness of deep learning models, with a particular focus on adversarial training. We aim to provide a comprehensive overall picture about this emerging direction and enable the community to be aware of the urgency and importance of designing robust deep learning models in safety-critical data analytical applications, ultimately enabling the end-users to trust deep learning classifiers. We will also summarize potential research directions concerning the adversarial robustness of deep learning, and its potential benefits to enable accountable and trustworthy deep learning-based data analytical systems and applications.
205
Human-Like Active Learning: Machines Simulating the Human Learning Process
Although the use of active learning to increase learners' engagement has recently been introduced in a variety of methods, empirical experiments are lacking. In this study, we attempted to align two experiments in order to (1) make a hypothesis for machine and (2) empirically confirm the effect of active learning on learning. In Experiment 1, we compared the effect of a passive form of learning to active form of learning. The results showed that active learning had a greater learning outcomes than passive learning. In the machine experiment based on the human result, we imitated the human active learning as a form of knowledge distillation. The active learning framework performed better than the passive learning framework. In the end, we showed not only that we can make build better machine training framework through the human experiment result, but also empirically confirm the result of human experiment through imitated machine experiments; human-like active learning have crucial effect on learning performance.
Disliked
zrz@andrew.cmu.edu
Human-Like Active Learning: Machines Simulating the Human Learning Process : Although the use of active learning to increase learners' engagement has recently been introduced in a variety of methods, empirical experiments are lacking. In this study, we attempted to align two experiments in order to (1) make a hypothesis for machine and (2) empirically confirm the effect of active learning on learning. In Experiment 1, we compared the effect of a passive form of learning to active form of learning. The results showed that active learning had a greater learning outcomes than passive learning. In the machine experiment based on the human result, we imitated the human active learning as a form of knowledge distillation. The active learning framework performed better than the passive learning framework. In the end, we showed not only that we can make build better machine training framework through the human experiment result, but also empirically confirm the result of human experiment through imitated machine experiments; human-like active learning have crucial effect on learning performance.
0
zrz@andrew.cmu.edu [SEP] Human-Like Active Learning: Machines Simulating the Human Learning Process : Although the use of active learning to increase learners' engagement has recently been introduced in a variety of methods, empirical experiments are lacking. In this study, we attempted to align two experiments in order to (1) make a hypothesis for machine and (2) empirically confirm the effect of active learning on learning. In Experiment 1, we compared the effect of a passive form of learning to active form of learning. The results showed that active learning had a greater learning outcomes than passive learning. In the machine experiment based on the human result, we imitated the human active learning as a form of knowledge distillation. The active learning framework performed better than the passive learning framework. In the end, we showed not only that we can make build better machine training framework through the human experiment result, but also empirically confirm the result of human experiment through imitated machine experiments; human-like active learning have crucial effect on learning performance.
128
Learning proofs for the classification of nilpotent semigroups
Machine learning is applied to find proofs, with smaller or smallest numbers of nodes, for the classification of 4-nilpotent semigroups.
Disliked
zrz@andrew.cmu.edu
Learning proofs for the classification of nilpotent semigroups : Machine learning is applied to find proofs, with smaller or smallest numbers of nodes, for the classification of 4-nilpotent semigroups.
0
zrz@andrew.cmu.edu [SEP] Learning proofs for the classification of nilpotent semigroups : Machine learning is applied to find proofs, with smaller or smallest numbers of nodes, for the classification of 4-nilpotent semigroups.
135
Flexible Morphing Aerial Robot with Inflatable Structure for Perching-based Human-Robot Interaction
Birds in nature perform perching not only for rest but also for interaction with human such as the relationship with falconers. Recently, researchers achieve perching-capable aerial robots as a way to save energy, and deformable structure demonstrate significant advantages in efficiency of perching and compactness of configuration. However, ensuring flight stability remains challenging for deformable aerial robots due to the difficulty of controlling flexible arms. Furthermore, perching for human interaction requires high compliance along with safety. Thus, this study aims to develop a deformable aerial robot capable of perching on humans with high flexibility and grasping ability. To overcome the challenges of stability of both flight and perching, we propose a hybrid morphing structure that combines a unilateral flexible arm and a pneumatic inflatable actuators. This design allows the robot's arms to remain rigid during flight and soft while perching for more effective grasping. We also develop a pneumatic control system that optimizes pressure regulation while integrating shock absorption and adjustable grasping forces, enhancing interaction capabilities and energy efficiency. Besides, we focus on the structural characteristics of the unilateral flexible arm and identify sufficient conditions under which standard quadrotor modeling and control remain effective in terms of flight stability. Finally, the developed prototype demonstrates the feasibility of compliant perching maneuvers on humans, as well as the robust recovery even after arm deformation caused by thrust reductions during flight. To the best of our knowledge, this work is the first to achieve an aerial robot capable of perching on humans for interaction.
Liked
jechoi@andrew.cmu.edu
Flexible Morphing Aerial Robot with Inflatable Structure for Perching-based Human-Robot Interaction : Birds in nature perform perching not only for rest but also for interaction with human such as the relationship with falconers. Recently, researchers achieve perching-capable aerial robots as a way to save energy, and deformable structure demonstrate significant advantages in efficiency of perching and compactness of configuration. However, ensuring flight stability remains challenging for deformable aerial robots due to the difficulty of controlling flexible arms. Furthermore, perching for human interaction requires high compliance along with safety. Thus, this study aims to develop a deformable aerial robot capable of perching on humans with high flexibility and grasping ability. To overcome the challenges of stability of both flight and perching, we propose a hybrid morphing structure that combines a unilateral flexible arm and a pneumatic inflatable actuators. This design allows the robot's arms to remain rigid during flight and soft while perching for more effective grasping. We also develop a pneumatic control system that optimizes pressure regulation while integrating shock absorption and adjustable grasping forces, enhancing interaction capabilities and energy efficiency. Besides, we focus on the structural characteristics of the unilateral flexible arm and identify sufficient conditions under which standard quadrotor modeling and control remain effective in terms of flight stability. Finally, the developed prototype demonstrates the feasibility of compliant perching maneuvers on humans, as well as the robust recovery even after arm deformation caused by thrust reductions during flight. To the best of our knowledge, this work is the first to achieve an aerial robot capable of perching on humans for interaction.
1
jechoi@andrew.cmu.edu [SEP] Flexible Morphing Aerial Robot with Inflatable Structure for Perching-based Human-Robot Interaction : Birds in nature perform perching not only for rest but also for interaction with human such as the relationship with falconers. Recently, researchers achieve perching-capable aerial robots as a way to save energy, and deformable structure demonstrate significant advantages in efficiency of perching and compactness of configuration. However, ensuring flight stability remains challenging for deformable aerial robots due to the difficulty of controlling flexible arms. Furthermore, perching for human interaction requires high compliance along with safety. Thus, this study aims to develop a deformable aerial robot capable of perching on humans with high flexibility and grasping ability. To overcome the challenges of stability of both flight and perching, we propose a hybrid morphing structure that combines a unilateral flexible arm and a pneumatic inflatable actuators. This design allows the robot's arms to remain rigid during flight and soft while perching for more effective grasping. We also develop a pneumatic control system that optimizes pressure regulation while integrating shock absorption and adjustable grasping forces, enhancing interaction capabilities and energy efficiency. Besides, we focus on the structural characteristics of the unilateral flexible arm and identify sufficient conditions under which standard quadrotor modeling and control remain effective in terms of flight stability. Finally, the developed prototype demonstrates the feasibility of compliant perching maneuvers on humans, as well as the robust recovery even after arm deformation caused by thrust reductions during flight. To the best of our knowledge, this work is the first to achieve an aerial robot capable of perching on humans for interaction.
525
Towards human-like kinematics in industrial robotic arms: a case study on a UR3 robot
Safety in industrial robotic environments is a hot research topic in the area of human-robot interaction (HRI). Up to now, a robotic arm on an assembly line interacts with other machines away from human workers. Nowadays, robotic arm manufactures are aimed to their robots could increasingly perform tasks collaborating with humans. One of the ways to improve this collaboration is by making the movement of robots more humanlike. This way, it would be easier for a human to foresee the movement of the robot and approach it without fear of contact. The main difference between the movement of a human and of a robotic arm is that the former has a bell-shaped speed profile while the latter has a uniform speed one. To generate this speed profile, the kinematic theory of rapid human movements and its Sigma-Lognormal model has been used. This model is widely used to explain most of the basic phenomena related to the control of human movements. Both human-like and robotic-like movements are transferred to the UR3 robot. In this paper we detail the how the UR3 robot was programmed to produce both kinds of movement. The dissimilarities result between the input motion and output motion to the robot confirm the possibility to develop human-like velocities in the UR3 robot.
Liked
jechoi@andrew.cmu.edu
Towards human-like kinematics in industrial robotic arms: a case study on a UR3 robot : Safety in industrial robotic environments is a hot research topic in the area of human-robot interaction (HRI). Up to now, a robotic arm on an assembly line interacts with other machines away from human workers. Nowadays, robotic arm manufactures are aimed to their robots could increasingly perform tasks collaborating with humans. One of the ways to improve this collaboration is by making the movement of robots more humanlike. This way, it would be easier for a human to foresee the movement of the robot and approach it without fear of contact. The main difference between the movement of a human and of a robotic arm is that the former has a bell-shaped speed profile while the latter has a uniform speed one. To generate this speed profile, the kinematic theory of rapid human movements and its Sigma-Lognormal model has been used. This model is widely used to explain most of the basic phenomena related to the control of human movements. Both human-like and robotic-like movements are transferred to the UR3 robot. In this paper we detail the how the UR3 robot was programmed to produce both kinds of movement. The dissimilarities result between the input motion and output motion to the robot confirm the possibility to develop human-like velocities in the UR3 robot.
1
jechoi@andrew.cmu.edu [SEP] Towards human-like kinematics in industrial robotic arms: a case study on a UR3 robot : Safety in industrial robotic environments is a hot research topic in the area of human-robot interaction (HRI). Up to now, a robotic arm on an assembly line interacts with other machines away from human workers. Nowadays, robotic arm manufactures are aimed to their robots could increasingly perform tasks collaborating with humans. One of the ways to improve this collaboration is by making the movement of robots more humanlike. This way, it would be easier for a human to foresee the movement of the robot and approach it without fear of contact. The main difference between the movement of a human and of a robotic arm is that the former has a bell-shaped speed profile while the latter has a uniform speed one. To generate this speed profile, the kinematic theory of rapid human movements and its Sigma-Lognormal model has been used. This model is widely used to explain most of the basic phenomena related to the control of human movements. Both human-like and robotic-like movements are transferred to the UR3 robot. In this paper we detail the how the UR3 robot was programmed to produce both kinds of movement. The dissimilarities result between the input motion and output motion to the robot confirm the possibility to develop human-like velocities in the UR3 robot.
440
An Optimal Control View of Adversarial Machine Learning
I describe an optimal control view of adversarial machine learning, where the dynamical system is the machine learner, the input are adversarial actions, and the control costs are defined by the adversary's goals to do harm and be hard to detect. This view encompasses many types of adversarial machine learning, including test-item attacks, training-data poisoning, and adversarial reward shaping. The view encourages adversarial machine learning researcher to utilize advances in control theory and reinforcement learning.
Disliked
zrz@andrew.cmu.edu
An Optimal Control View of Adversarial Machine Learning : I describe an optimal control view of adversarial machine learning, where the dynamical system is the machine learner, the input are adversarial actions, and the control costs are defined by the adversary's goals to do harm and be hard to detect. This view encompasses many types of adversarial machine learning, including test-item attacks, training-data poisoning, and adversarial reward shaping. The view encourages adversarial machine learning researcher to utilize advances in control theory and reinforcement learning.
0
zrz@andrew.cmu.edu [SEP] An Optimal Control View of Adversarial Machine Learning : I describe an optimal control view of adversarial machine learning, where the dynamical system is the machine learner, the input are adversarial actions, and the control costs are defined by the adversary's goals to do harm and be hard to detect. This view encompasses many types of adversarial machine learning, including test-item attacks, training-data poisoning, and adversarial reward shaping. The view encourages adversarial machine learning researcher to utilize advances in control theory and reinforcement learning.
27
Analyzing Fine-Grained Alignment and Enhancing Vision Understanding in Multimodal Language Models
Achieving better alignment between vision embeddings and Large Language Models (LLMs) is crucial for enhancing the abilities of Multimodal LLMs (MLLMs), particularly for recent models that rely on powerful pretrained vision encoders and LLMs. A common approach to connect the pretrained vision encoder and LLM is through a projector applied after the vision encoder. However, the projector is often trained to enable the LLM to generate captions, and hence the mechanism by which LLMs understand each vision token remains unclear. In this work, we first investigate the role of the projector in compressing vision embeddings and aligning them with word embeddings. We show that the projector significantly compresses visual information, removing redundant details while preserving essential elements necessary for the LLM to understand visual content. We then examine patch-level alignment -- the alignment between each vision patch and its corresponding semantic words -- and propose a *multi-semantic alignment hypothesis*. Our analysis indicates that the projector trained by caption loss improves patch-level alignment but only to a limited extent, resulting in weak and coarse alignment. To address this issue, we propose *patch-aligned training* to efficiently enhance patch-level alignment. Our experiments show that patch-aligned training (1) achieves stronger compression capability and improved patch-level alignment, enabling the MLLM to generate higher-quality captions, (2) improves the MLLM's performance by 16% on referring expression grounding tasks, 4% on question-answering tasks, and 3% on modern instruction-following benchmarks when using the same supervised fine-tuning (SFT) setting. The proposed method can be easily extended to other multimodal models.
Liked
zrz@andrew.cmu.edu
Analyzing Fine-Grained Alignment and Enhancing Vision Understanding in Multimodal Language Models : Achieving better alignment between vision embeddings and Large Language Models (LLMs) is crucial for enhancing the abilities of Multimodal LLMs (MLLMs), particularly for recent models that rely on powerful pretrained vision encoders and LLMs. A common approach to connect the pretrained vision encoder and LLM is through a projector applied after the vision encoder. However, the projector is often trained to enable the LLM to generate captions, and hence the mechanism by which LLMs understand each vision token remains unclear. In this work, we first investigate the role of the projector in compressing vision embeddings and aligning them with word embeddings. We show that the projector significantly compresses visual information, removing redundant details while preserving essential elements necessary for the LLM to understand visual content. We then examine patch-level alignment -- the alignment between each vision patch and its corresponding semantic words -- and propose a *multi-semantic alignment hypothesis*. Our analysis indicates that the projector trained by caption loss improves patch-level alignment but only to a limited extent, resulting in weak and coarse alignment. To address this issue, we propose *patch-aligned training* to efficiently enhance patch-level alignment. Our experiments show that patch-aligned training (1) achieves stronger compression capability and improved patch-level alignment, enabling the MLLM to generate higher-quality captions, (2) improves the MLLM's performance by 16% on referring expression grounding tasks, 4% on question-answering tasks, and 3% on modern instruction-following benchmarks when using the same supervised fine-tuning (SFT) setting. The proposed method can be easily extended to other multimodal models.
1
zrz@andrew.cmu.edu [SEP] Analyzing Fine-Grained Alignment and Enhancing Vision Understanding in Multimodal Language Models : Achieving better alignment between vision embeddings and Large Language Models (LLMs) is crucial for enhancing the abilities of Multimodal LLMs (MLLMs), particularly for recent models that rely on powerful pretrained vision encoders and LLMs. A common approach to connect the pretrained vision encoder and LLM is through a projector applied after the vision encoder. However, the projector is often trained to enable the LLM to generate captions, and hence the mechanism by which LLMs understand each vision token remains unclear. In this work, we first investigate the role of the projector in compressing vision embeddings and aligning them with word embeddings. We show that the projector significantly compresses visual information, removing redundant details while preserving essential elements necessary for the LLM to understand visual content. We then examine patch-level alignment -- the alignment between each vision patch and its corresponding semantic words -- and propose a *multi-semantic alignment hypothesis*. Our analysis indicates that the projector trained by caption loss improves patch-level alignment but only to a limited extent, resulting in weak and coarse alignment. To address this issue, we propose *patch-aligned training* to efficiently enhance patch-level alignment. Our experiments show that patch-aligned training (1) achieves stronger compression capability and improved patch-level alignment, enabling the MLLM to generate higher-quality captions, (2) improves the MLLM's performance by 16% on referring expression grounding tasks, 4% on question-answering tasks, and 3% on modern instruction-following benchmarks when using the same supervised fine-tuning (SFT) setting. The proposed method can be easily extended to other multimodal models.
376
Learning Human-arm Reaching Motion Using IMU in Human-Robot Collaboration
Many tasks performed by two humans require mutual interaction between arms such as handing-over tools and objects. In order for a robotic arm to interact with a human in the same way, it must reason about the location of the human arm in real-time. Furthermore and to acquire interaction in a timely manner, the robot must be able predict the final target of the human in order to plan and initiate motion beforehand. In this paper, we explore the use of a low-cost wearable device equipped with two inertial measurement units (IMU) for learning reaching motion for real-time applications of Human-Robot Collaboration (HRC). A wearable device can replace or be complementary to visual perception in cases of bad lighting or occlusions in a cluttered environment. We first train a neural-network model to estimate the current location of the arm. Then, we propose a novel model based on a recurrent neural-network to predict the future target of the human arm during motion in real-time. Early prediction of the target grants the robot with sufficient time to plan and initiate motion during the motion of the human. The accuracies of the models are analyzed concerning the features included in the motion representation. Through experiments and real demonstrations with a robotic arm, we show that sufficient accuracy is achieved for feasible HRC without any visual perception. Once trained, the system can be deployed in various spaces with no additional effort. The models exhibit high accuracy for various initial poses of the human arm. Moreover, the trained models are shown to provide high success rates with additional human participants not included in the model training.
Liked
jechoi@andrew.cmu.edu
Learning Human-arm Reaching Motion Using IMU in Human-Robot Collaboration : Many tasks performed by two humans require mutual interaction between arms such as handing-over tools and objects. In order for a robotic arm to interact with a human in the same way, it must reason about the location of the human arm in real-time. Furthermore and to acquire interaction in a timely manner, the robot must be able predict the final target of the human in order to plan and initiate motion beforehand. In this paper, we explore the use of a low-cost wearable device equipped with two inertial measurement units (IMU) for learning reaching motion for real-time applications of Human-Robot Collaboration (HRC). A wearable device can replace or be complementary to visual perception in cases of bad lighting or occlusions in a cluttered environment. We first train a neural-network model to estimate the current location of the arm. Then, we propose a novel model based on a recurrent neural-network to predict the future target of the human arm during motion in real-time. Early prediction of the target grants the robot with sufficient time to plan and initiate motion during the motion of the human. The accuracies of the models are analyzed concerning the features included in the motion representation. Through experiments and real demonstrations with a robotic arm, we show that sufficient accuracy is achieved for feasible HRC without any visual perception. Once trained, the system can be deployed in various spaces with no additional effort. The models exhibit high accuracy for various initial poses of the human arm. Moreover, the trained models are shown to provide high success rates with additional human participants not included in the model training.
1
jechoi@andrew.cmu.edu [SEP] Learning Human-arm Reaching Motion Using IMU in Human-Robot Collaboration : Many tasks performed by two humans require mutual interaction between arms such as handing-over tools and objects. In order for a robotic arm to interact with a human in the same way, it must reason about the location of the human arm in real-time. Furthermore and to acquire interaction in a timely manner, the robot must be able predict the final target of the human in order to plan and initiate motion beforehand. In this paper, we explore the use of a low-cost wearable device equipped with two inertial measurement units (IMU) for learning reaching motion for real-time applications of Human-Robot Collaboration (HRC). A wearable device can replace or be complementary to visual perception in cases of bad lighting or occlusions in a cluttered environment. We first train a neural-network model to estimate the current location of the arm. Then, we propose a novel model based on a recurrent neural-network to predict the future target of the human arm during motion in real-time. Early prediction of the target grants the robot with sufficient time to plan and initiate motion during the motion of the human. The accuracies of the models are analyzed concerning the features included in the motion representation. Through experiments and real demonstrations with a robotic arm, we show that sufficient accuracy is achieved for feasible HRC without any visual perception. Once trained, the system can be deployed in various spaces with no additional effort. The models exhibit high accuracy for various initial poses of the human arm. Moreover, the trained models are shown to provide high success rates with additional human participants not included in the model training.
441
Adding Context to Source Code Representations for Deep Learning
Deep learning models have been successfully applied to a variety of software engineering tasks, such as code classification, summarisation, and bug and vulnerability detection. In order to apply deep learning to these tasks, source code needs to be represented in a format that is suitable for input into the deep learning model. Most approaches to representing source code, such as tokens, abstract syntax trees (ASTs), data flow graphs (DFGs), and control flow graphs (CFGs) only focus on the code itself and do not take into account additional context that could be useful for deep learning models. In this paper, we argue that it is beneficial for deep learning models to have access to additional contextual information about the code being analysed. We present preliminary evidence that encoding context from the call hierarchy along with information from the code itself can improve the performance of a state-of-the-art deep learning model for two software engineering tasks. We outline our research agenda for adding further contextual information to source code representations for deep learning.
Liked
zrz@andrew.cmu.edu
Adding Context to Source Code Representations for Deep Learning : Deep learning models have been successfully applied to a variety of software engineering tasks, such as code classification, summarisation, and bug and vulnerability detection. In order to apply deep learning to these tasks, source code needs to be represented in a format that is suitable for input into the deep learning model. Most approaches to representing source code, such as tokens, abstract syntax trees (ASTs), data flow graphs (DFGs), and control flow graphs (CFGs) only focus on the code itself and do not take into account additional context that could be useful for deep learning models. In this paper, we argue that it is beneficial for deep learning models to have access to additional contextual information about the code being analysed. We present preliminary evidence that encoding context from the call hierarchy along with information from the code itself can improve the performance of a state-of-the-art deep learning model for two software engineering tasks. We outline our research agenda for adding further contextual information to source code representations for deep learning.
1
zrz@andrew.cmu.edu [SEP] Adding Context to Source Code Representations for Deep Learning : Deep learning models have been successfully applied to a variety of software engineering tasks, such as code classification, summarisation, and bug and vulnerability detection. In order to apply deep learning to these tasks, source code needs to be represented in a format that is suitable for input into the deep learning model. Most approaches to representing source code, such as tokens, abstract syntax trees (ASTs), data flow graphs (DFGs), and control flow graphs (CFGs) only focus on the code itself and do not take into account additional context that could be useful for deep learning models. In this paper, we argue that it is beneficial for deep learning models to have access to additional contextual information about the code being analysed. We present preliminary evidence that encoding context from the call hierarchy along with information from the code itself can improve the performance of a state-of-the-art deep learning model for two software engineering tasks. We outline our research agenda for adding further contextual information to source code representations for deep learning.
246
Multi-Modal Masked Autoencoders for Medical Vision-and-Language Pre-Training
Medical vision-and-language pre-training provides a feasible solution to extract effective vision-and-language representations from medical images and texts. However, few studies have been dedicated to this field to facilitate medical vision-and-language understanding. In this paper, we propose a self-supervised learning paradigm with multi-modal masked autoencoders (M$^3$AE), which learn cross-modal domain knowledge by reconstructing missing pixels and tokens from randomly masked images and texts. There are three key designs to make this simple approach work. First, considering the different information densities of vision and language, we adopt different masking ratios for the input image and text, where a considerably larger masking ratio is used for images. Second, we use visual and textual features from different layers to perform the reconstruction to deal with different levels of abstraction in visual and language. Third, we develop different designs for vision and language decoders (i.e., a Transformer for vision and a multi-layer perceptron for language). To perform a comprehensive evaluation and facilitate further research, we construct a medical vision-and-language benchmark including three tasks. Experimental results demonstrate the effectiveness of our approach, where state-of-the-art results are achieved on all downstream tasks. Besides, we conduct further analysis to better verify the effectiveness of different components of our approach and various settings of pre-training. The source code is available at~\url{https://github.com/zhjohnchan/M3AE}.
Liked
zrz@andrew.cmu.edu
Multi-Modal Masked Autoencoders for Medical Vision-and-Language Pre-Training : Medical vision-and-language pre-training provides a feasible solution to extract effective vision-and-language representations from medical images and texts. However, few studies have been dedicated to this field to facilitate medical vision-and-language understanding. In this paper, we propose a self-supervised learning paradigm with multi-modal masked autoencoders (M$^3$AE), which learn cross-modal domain knowledge by reconstructing missing pixels and tokens from randomly masked images and texts. There are three key designs to make this simple approach work. First, considering the different information densities of vision and language, we adopt different masking ratios for the input image and text, where a considerably larger masking ratio is used for images. Second, we use visual and textual features from different layers to perform the reconstruction to deal with different levels of abstraction in visual and language. Third, we develop different designs for vision and language decoders (i.e., a Transformer for vision and a multi-layer perceptron for language). To perform a comprehensive evaluation and facilitate further research, we construct a medical vision-and-language benchmark including three tasks. Experimental results demonstrate the effectiveness of our approach, where state-of-the-art results are achieved on all downstream tasks. Besides, we conduct further analysis to better verify the effectiveness of different components of our approach and various settings of pre-training. The source code is available at~\url{https://github.com/zhjohnchan/M3AE}.
1
zrz@andrew.cmu.edu [SEP] Multi-Modal Masked Autoencoders for Medical Vision-and-Language Pre-Training : Medical vision-and-language pre-training provides a feasible solution to extract effective vision-and-language representations from medical images and texts. However, few studies have been dedicated to this field to facilitate medical vision-and-language understanding. In this paper, we propose a self-supervised learning paradigm with multi-modal masked autoencoders (M$^3$AE), which learn cross-modal domain knowledge by reconstructing missing pixels and tokens from randomly masked images and texts. There are three key designs to make this simple approach work. First, considering the different information densities of vision and language, we adopt different masking ratios for the input image and text, where a considerably larger masking ratio is used for images. Second, we use visual and textual features from different layers to perform the reconstruction to deal with different levels of abstraction in visual and language. Third, we develop different designs for vision and language decoders (i.e., a Transformer for vision and a multi-layer perceptron for language). To perform a comprehensive evaluation and facilitate further research, we construct a medical vision-and-language benchmark including three tasks. Experimental results demonstrate the effectiveness of our approach, where state-of-the-art results are achieved on all downstream tasks. Besides, we conduct further analysis to better verify the effectiveness of different components of our approach and various settings of pre-training. The source code is available at~\url{https://github.com/zhjohnchan/M3AE}.
367
Modern Deep Reinforcement Learning Algorithms
Recent advances in Reinforcement Learning, grounded on combining classical theoretical results with Deep Learning paradigm, led to breakthroughs in many artificial intelligence tasks and gave birth to Deep Reinforcement Learning (DRL) as a field of research. In this work latest DRL algorithms are reviewed with a focus on their theoretical justification, practical limitations and observed empirical properties.
Liked
zrz@andrew.cmu.edu
Modern Deep Reinforcement Learning Algorithms : Recent advances in Reinforcement Learning, grounded on combining classical theoretical results with Deep Learning paradigm, led to breakthroughs in many artificial intelligence tasks and gave birth to Deep Reinforcement Learning (DRL) as a field of research. In this work latest DRL algorithms are reviewed with a focus on their theoretical justification, practical limitations and observed empirical properties.
1
zrz@andrew.cmu.edu [SEP] Modern Deep Reinforcement Learning Algorithms : Recent advances in Reinforcement Learning, grounded on combining classical theoretical results with Deep Learning paradigm, led to breakthroughs in many artificial intelligence tasks and gave birth to Deep Reinforcement Learning (DRL) as a field of research. In this work latest DRL algorithms are reviewed with a focus on their theoretical justification, practical limitations and observed empirical properties.
212
An Introduction to MM Algorithms for Machine Learning and Statistical
MM (majorization--minimization) algorithms are an increasingly popular tool for solving optimization problems in machine learning and statistical estimation. This article introduces the MM algorithm framework in general and via three popular example applications: Gaussian mixture regressions, multinomial logistic regressions, and support vector machines. Specific algorithms for the three examples are derived and numerical demonstrations are presented. Theoretical and practical aspects of MM algorithm design are discussed.
Liked
zrz@andrew.cmu.edu
An Introduction to MM Algorithms for Machine Learning and Statistical : MM (majorization--minimization) algorithms are an increasingly popular tool for solving optimization problems in machine learning and statistical estimation. This article introduces the MM algorithm framework in general and via three popular example applications: Gaussian mixture regressions, multinomial logistic regressions, and support vector machines. Specific algorithms for the three examples are derived and numerical demonstrations are presented. Theoretical and practical aspects of MM algorithm design are discussed.
1
zrz@andrew.cmu.edu [SEP] An Introduction to MM Algorithms for Machine Learning and Statistical : MM (majorization--minimization) algorithms are an increasingly popular tool for solving optimization problems in machine learning and statistical estimation. This article introduces the MM algorithm framework in general and via three popular example applications: Gaussian mixture regressions, multinomial logistic regressions, and support vector machines. Specific algorithms for the three examples are derived and numerical demonstrations are presented. Theoretical and practical aspects of MM algorithm design are discussed.
97
The SET Perceptual Factors Framework: Towards Assured Perception for Autonomous Systems
Future autonomous systems promise significant societal benefits, yet their deployment raises concerns about safety and trustworthiness. A key concern is assuring the reliability of robot perception, as perception seeds safe decision-making. Failures in perception are often due to complex yet common environmental factors and can lead to accidents that erode public trust. To address this concern, we introduce the SET (Self, Environment, and Target) Perceptual Factors Framework. We designed the framework to systematically analyze how factors such as weather, occlusion, or sensor limitations negatively impact perception. To achieve this, the framework employs SET State Trees to categorize where such factors originate and SET Factor Trees to model how these sources and factors impact perceptual tasks like object detection or pose estimation. Next, we develop Perceptual Factor Models using both trees to quantify the uncertainty for a given task. Our framework aims to promote rigorous safety assurances and cultivate greater public understanding and trust in autonomous systems by offering a transparent and standardized method for identifying, modeling, and communicating perceptual risks.
Disliked
zrz@andrew.cmu.edu
The SET Perceptual Factors Framework: Towards Assured Perception for Autonomous Systems : Future autonomous systems promise significant societal benefits, yet their deployment raises concerns about safety and trustworthiness. A key concern is assuring the reliability of robot perception, as perception seeds safe decision-making. Failures in perception are often due to complex yet common environmental factors and can lead to accidents that erode public trust. To address this concern, we introduce the SET (Self, Environment, and Target) Perceptual Factors Framework. We designed the framework to systematically analyze how factors such as weather, occlusion, or sensor limitations negatively impact perception. To achieve this, the framework employs SET State Trees to categorize where such factors originate and SET Factor Trees to model how these sources and factors impact perceptual tasks like object detection or pose estimation. Next, we develop Perceptual Factor Models using both trees to quantify the uncertainty for a given task. Our framework aims to promote rigorous safety assurances and cultivate greater public understanding and trust in autonomous systems by offering a transparent and standardized method for identifying, modeling, and communicating perceptual risks.
0
zrz@andrew.cmu.edu [SEP] The SET Perceptual Factors Framework: Towards Assured Perception for Autonomous Systems : Future autonomous systems promise significant societal benefits, yet their deployment raises concerns about safety and trustworthiness. A key concern is assuring the reliability of robot perception, as perception seeds safe decision-making. Failures in perception are often due to complex yet common environmental factors and can lead to accidents that erode public trust. To address this concern, we introduce the SET (Self, Environment, and Target) Perceptual Factors Framework. We designed the framework to systematically analyze how factors such as weather, occlusion, or sensor limitations negatively impact perception. To achieve this, the framework employs SET State Trees to categorize where such factors originate and SET Factor Trees to model how these sources and factors impact perceptual tasks like object detection or pose estimation. Next, we develop Perceptual Factor Models using both trees to quantify the uncertainty for a given task. Our framework aims to promote rigorous safety assurances and cultivate greater public understanding and trust in autonomous systems by offering a transparent and standardized method for identifying, modeling, and communicating perceptual risks.
299
Emotional Musical Prosody for the Enhancement of Trust in Robotic Arm Communication
As robotic arms become prevalent in industry it is crucial to improve levels of trust from human collaborators. Low levels of trust in human-robot interaction can reduce overall performance and prevent full robot utilization. We investigated the potential benefits of using emotional musical prosody to allow the robot to respond emotionally to the user's actions. We tested participants' responses to interacting with a virtual robot arm that acted as a decision agent, helping participants select the next number in a sequence. We compared results from three versions of the application in a between-group experiment, where the robot had different emotional reactions to the user's input depending on whether the user agreed with the robot and whether the user's choice was correct. In all versions, the robot reacted with emotional gestures. One version used prosody-based emotional audio phrases selected from our dataset of singer improvisations, the second version used audio consisting of a single pitch randomly assigned to each emotion, and the final version used no audio, only gestures. Our results showed no significant difference for the percentage of times users from each group agreed with the robot, and no difference between user's agreement with the robot after it made a mistake. However, participants also took a trust survey following the interaction, and we found that the reported trust ratings of the musical prosody group were significantly higher than both the single-pitch and no audio groups.
Disliked
jechoi@andrew.cmu.edu
Emotional Musical Prosody for the Enhancement of Trust in Robotic Arm Communication : As robotic arms become prevalent in industry it is crucial to improve levels of trust from human collaborators. Low levels of trust in human-robot interaction can reduce overall performance and prevent full robot utilization. We investigated the potential benefits of using emotional musical prosody to allow the robot to respond emotionally to the user's actions. We tested participants' responses to interacting with a virtual robot arm that acted as a decision agent, helping participants select the next number in a sequence. We compared results from three versions of the application in a between-group experiment, where the robot had different emotional reactions to the user's input depending on whether the user agreed with the robot and whether the user's choice was correct. In all versions, the robot reacted with emotional gestures. One version used prosody-based emotional audio phrases selected from our dataset of singer improvisations, the second version used audio consisting of a single pitch randomly assigned to each emotion, and the final version used no audio, only gestures. Our results showed no significant difference for the percentage of times users from each group agreed with the robot, and no difference between user's agreement with the robot after it made a mistake. However, participants also took a trust survey following the interaction, and we found that the reported trust ratings of the musical prosody group were significantly higher than both the single-pitch and no audio groups.
0
jechoi@andrew.cmu.edu [SEP] Emotional Musical Prosody for the Enhancement of Trust in Robotic Arm Communication : As robotic arms become prevalent in industry it is crucial to improve levels of trust from human collaborators. Low levels of trust in human-robot interaction can reduce overall performance and prevent full robot utilization. We investigated the potential benefits of using emotional musical prosody to allow the robot to respond emotionally to the user's actions. We tested participants' responses to interacting with a virtual robot arm that acted as a decision agent, helping participants select the next number in a sequence. We compared results from three versions of the application in a between-group experiment, where the robot had different emotional reactions to the user's input depending on whether the user agreed with the robot and whether the user's choice was correct. In all versions, the robot reacted with emotional gestures. One version used prosody-based emotional audio phrases selected from our dataset of singer improvisations, the second version used audio consisting of a single pitch randomly assigned to each emotion, and the final version used no audio, only gestures. Our results showed no significant difference for the percentage of times users from each group agreed with the robot, and no difference between user's agreement with the robot after it made a mistake. However, participants also took a trust survey following the interaction, and we found that the reported trust ratings of the musical prosody group were significantly higher than both the single-pitch and no audio groups.
542
Generating Realistic Arm Movements in Reinforcement Learning: A Quantitative Comparison of Reward Terms and Task Requirements
The mimicking of human-like arm movement characteristics involves the consideration of three factors during control policy synthesis: (a) chosen task requirements, (b) inclusion of noise during movement execution and (c) chosen optimality principles. Previous studies showed that when considering these factors (a-c) individually, it is possible to synthesize arm movements that either kinematically match the experimental data or reproduce the stereotypical triphasic muscle activation pattern. However, to date no quantitative comparison has been made on how realistic the arm movement generated by each factor is; as well as whether a partial or total combination of all factors results in arm movements with human-like kinematic characteristics and a triphasic muscle pattern. To investigate this, we used reinforcement learning to learn a control policy for a musculoskeletal arm model, aiming to discern which combination of factors (a-c) results in realistic arm movements according to four frequently reported stereotypical characteristics. Our findings indicate that incorporating velocity and acceleration requirements into the reaching task, employing reward terms that encourage minimization of mechanical work, hand jerk, and control effort, along with the inclusion of noise during movement, leads to the emergence of realistic human arm movements in reinforcement learning. We expect that the gained insights will help in the future to better predict desired arm movements and corrective forces in wearable assistive devices.
Liked
jechoi@andrew.cmu.edu
Generating Realistic Arm Movements in Reinforcement Learning: A Quantitative Comparison of Reward Terms and Task Requirements : The mimicking of human-like arm movement characteristics involves the consideration of three factors during control policy synthesis: (a) chosen task requirements, (b) inclusion of noise during movement execution and (c) chosen optimality principles. Previous studies showed that when considering these factors (a-c) individually, it is possible to synthesize arm movements that either kinematically match the experimental data or reproduce the stereotypical triphasic muscle activation pattern. However, to date no quantitative comparison has been made on how realistic the arm movement generated by each factor is; as well as whether a partial or total combination of all factors results in arm movements with human-like kinematic characteristics and a triphasic muscle pattern. To investigate this, we used reinforcement learning to learn a control policy for a musculoskeletal arm model, aiming to discern which combination of factors (a-c) results in realistic arm movements according to four frequently reported stereotypical characteristics. Our findings indicate that incorporating velocity and acceleration requirements into the reaching task, employing reward terms that encourage minimization of mechanical work, hand jerk, and control effort, along with the inclusion of noise during movement, leads to the emergence of realistic human arm movements in reinforcement learning. We expect that the gained insights will help in the future to better predict desired arm movements and corrective forces in wearable assistive devices.
1
jechoi@andrew.cmu.edu [SEP] Generating Realistic Arm Movements in Reinforcement Learning: A Quantitative Comparison of Reward Terms and Task Requirements : The mimicking of human-like arm movement characteristics involves the consideration of three factors during control policy synthesis: (a) chosen task requirements, (b) inclusion of noise during movement execution and (c) chosen optimality principles. Previous studies showed that when considering these factors (a-c) individually, it is possible to synthesize arm movements that either kinematically match the experimental data or reproduce the stereotypical triphasic muscle activation pattern. However, to date no quantitative comparison has been made on how realistic the arm movement generated by each factor is; as well as whether a partial or total combination of all factors results in arm movements with human-like kinematic characteristics and a triphasic muscle pattern. To investigate this, we used reinforcement learning to learn a control policy for a musculoskeletal arm model, aiming to discern which combination of factors (a-c) results in realistic arm movements according to four frequently reported stereotypical characteristics. Our findings indicate that incorporating velocity and acceleration requirements into the reaching task, employing reward terms that encourage minimization of mechanical work, hand jerk, and control effort, along with the inclusion of noise during movement, leads to the emergence of realistic human arm movements in reinforcement learning. We expect that the gained insights will help in the future to better predict desired arm movements and corrective forces in wearable assistive devices.
571
A Unified Analytical Framework for Trustable Machine Learning and Automation Running with Blockchain
Traditional machine learning algorithms use data from databases that are mutable, and therefore the data cannot be fully trusted. Also, the machine learning process is difficult to automate. This paper proposes building a trustable machine learning system by using blockchain technology, which can store data in a permanent and immutable way. In addition, smart contracts are used to automate the machine learning process. This paper makes three contributions. First, it establishes a link between machine learning technology and blockchain technology. Previously, machine learning and blockchain have been considered two independent technologies without an obvious link. Second, it proposes a unified analytical framework for trustable machine learning by using blockchain technology. This unified framework solves both the trustability and automation issues in machine learning. Third, it enables a computer to translate core machine learning implementation from a single thread on a single machine to multiple threads on multiple machines running with blockchain by using a unified approach. The paper uses association rule mining as an example to demonstrate how trustable machine learning can be implemented with blockchain, and it shows how this approach can be used to analyze opioid prescriptions to help combat the opioid crisis.
Disliked
zrz@andrew.cmu.edu
A Unified Analytical Framework for Trustable Machine Learning and Automation Running with Blockchain : Traditional machine learning algorithms use data from databases that are mutable, and therefore the data cannot be fully trusted. Also, the machine learning process is difficult to automate. This paper proposes building a trustable machine learning system by using blockchain technology, which can store data in a permanent and immutable way. In addition, smart contracts are used to automate the machine learning process. This paper makes three contributions. First, it establishes a link between machine learning technology and blockchain technology. Previously, machine learning and blockchain have been considered two independent technologies without an obvious link. Second, it proposes a unified analytical framework for trustable machine learning by using blockchain technology. This unified framework solves both the trustability and automation issues in machine learning. Third, it enables a computer to translate core machine learning implementation from a single thread on a single machine to multiple threads on multiple machines running with blockchain by using a unified approach. The paper uses association rule mining as an example to demonstrate how trustable machine learning can be implemented with blockchain, and it shows how this approach can be used to analyze opioid prescriptions to help combat the opioid crisis.
0
zrz@andrew.cmu.edu [SEP] A Unified Analytical Framework for Trustable Machine Learning and Automation Running with Blockchain : Traditional machine learning algorithms use data from databases that are mutable, and therefore the data cannot be fully trusted. Also, the machine learning process is difficult to automate. This paper proposes building a trustable machine learning system by using blockchain technology, which can store data in a permanent and immutable way. In addition, smart contracts are used to automate the machine learning process. This paper makes three contributions. First, it establishes a link between machine learning technology and blockchain technology. Previously, machine learning and blockchain have been considered two independent technologies without an obvious link. Second, it proposes a unified analytical framework for trustable machine learning by using blockchain technology. This unified framework solves both the trustability and automation issues in machine learning. Third, it enables a computer to translate core machine learning implementation from a single thread on a single machine to multiple threads on multiple machines running with blockchain by using a unified approach. The paper uses association rule mining as an example to demonstrate how trustable machine learning can be implemented with blockchain, and it shows how this approach can be used to analyze opioid prescriptions to help combat the opioid crisis.
35
A Survey of Deep Learning Techniques for Mobile Robot Applications
Advancements in deep learning over the years have attracted research into how deep artificial neural networks can be used in robotic systems. This research survey will present a summarization of the current research with a specific focus on the gains and obstacles for deep learning to be applied to mobile robotics.
Liked
zrz@andrew.cmu.edu
A Survey of Deep Learning Techniques for Mobile Robot Applications : Advancements in deep learning over the years have attracted research into how deep artificial neural networks can be used in robotic systems. This research survey will present a summarization of the current research with a specific focus on the gains and obstacles for deep learning to be applied to mobile robotics.
1
zrz@andrew.cmu.edu [SEP] A Survey of Deep Learning Techniques for Mobile Robot Applications : Advancements in deep learning over the years have attracted research into how deep artificial neural networks can be used in robotic systems. This research survey will present a summarization of the current research with a specific focus on the gains and obstacles for deep learning to be applied to mobile robotics.
267
Moving Deep Learning into Web Browser: How Far Can We Go?
Recently, several JavaScript-based deep learning frameworks have emerged, making it possible to perform deep learning tasks directly in browsers. However, little is known on what and how well we can do with these frameworks for deep learning in browsers. To bridge the knowledge gap, in this paper, we conduct the first empirical study of deep learning in browsers. We survey 7 most popular JavaScript-based deep learning frameworks, investigating to what extent deep learning tasks have been supported in browsers so far. Then we measure the performance of different frameworks when running different deep learning tasks. Finally, we dig out the performance gap between deep learning in browsers and on native platforms by comparing the performance of TensorFlow.js and TensorFlow in Python. Our findings could help application developers, deep-learning framework vendors and browser vendors to improve the efficiency of deep learning in browsers.
Disliked
zrz@andrew.cmu.edu
Moving Deep Learning into Web Browser: How Far Can We Go? : Recently, several JavaScript-based deep learning frameworks have emerged, making it possible to perform deep learning tasks directly in browsers. However, little is known on what and how well we can do with these frameworks for deep learning in browsers. To bridge the knowledge gap, in this paper, we conduct the first empirical study of deep learning in browsers. We survey 7 most popular JavaScript-based deep learning frameworks, investigating to what extent deep learning tasks have been supported in browsers so far. Then we measure the performance of different frameworks when running different deep learning tasks. Finally, we dig out the performance gap between deep learning in browsers and on native platforms by comparing the performance of TensorFlow.js and TensorFlow in Python. Our findings could help application developers, deep-learning framework vendors and browser vendors to improve the efficiency of deep learning in browsers.
0
zrz@andrew.cmu.edu [SEP] Moving Deep Learning into Web Browser: How Far Can We Go? : Recently, several JavaScript-based deep learning frameworks have emerged, making it possible to perform deep learning tasks directly in browsers. However, little is known on what and how well we can do with these frameworks for deep learning in browsers. To bridge the knowledge gap, in this paper, we conduct the first empirical study of deep learning in browsers. We survey 7 most popular JavaScript-based deep learning frameworks, investigating to what extent deep learning tasks have been supported in browsers so far. Then we measure the performance of different frameworks when running different deep learning tasks. Finally, we dig out the performance gap between deep learning in browsers and on native platforms by comparing the performance of TensorFlow.js and TensorFlow in Python. Our findings could help application developers, deep-learning framework vendors and browser vendors to improve the efficiency of deep learning in browsers.
164
Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning
Deep reinforcement learning (deep RL) has achieved superior performance in complex sequential tasks by using deep neural networks as function approximators to learn directly from raw input images. However, learning directly from raw images is data inefficient. The agent must learn feature representation of complex states in addition to learning a policy. As a result, deep RL typically suffers from slow learning speeds and often requires a prohibitively large amount of training time and data to reach reasonable performance, making it inapplicable to real-world settings where data is expensive. In this work, we improve data efficiency in deep RL by addressing one of the two learning goals, feature learning. We leverage supervised learning to pre-train on a small set of non-expert human demonstrations and empirically evaluate our approach using the asynchronous advantage actor-critic algorithms (A3C) in the Atari domain. Our results show significant improvements in learning speed, even when the provided demonstration is noisy and of low quality.
Liked
zrz@andrew.cmu.edu
Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning : Deep reinforcement learning (deep RL) has achieved superior performance in complex sequential tasks by using deep neural networks as function approximators to learn directly from raw input images. However, learning directly from raw images is data inefficient. The agent must learn feature representation of complex states in addition to learning a policy. As a result, deep RL typically suffers from slow learning speeds and often requires a prohibitively large amount of training time and data to reach reasonable performance, making it inapplicable to real-world settings where data is expensive. In this work, we improve data efficiency in deep RL by addressing one of the two learning goals, feature learning. We leverage supervised learning to pre-train on a small set of non-expert human demonstrations and empirically evaluate our approach using the asynchronous advantage actor-critic algorithms (A3C) in the Atari domain. Our results show significant improvements in learning speed, even when the provided demonstration is noisy and of low quality.
1
zrz@andrew.cmu.edu [SEP] Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning : Deep reinforcement learning (deep RL) has achieved superior performance in complex sequential tasks by using deep neural networks as function approximators to learn directly from raw input images. However, learning directly from raw images is data inefficient. The agent must learn feature representation of complex states in addition to learning a policy. As a result, deep RL typically suffers from slow learning speeds and often requires a prohibitively large amount of training time and data to reach reasonable performance, making it inapplicable to real-world settings where data is expensive. In this work, we improve data efficiency in deep RL by addressing one of the two learning goals, feature learning. We leverage supervised learning to pre-train on a small set of non-expert human demonstrations and empirically evaluate our approach using the asynchronous advantage actor-critic algorithms (A3C) in the Atari domain. Our results show significant improvements in learning speed, even when the provided demonstration is noisy and of low quality.
260
Beyond One Model Fits All: Ensemble Deep Learning for Autonomous Vehicles
Deep learning has revolutionized autonomous driving by enabling vehicles to perceive and interpret their surroundings with remarkable accuracy. This progress is attributed to various deep learning models, including Mediated Perception, Behavior Reflex, and Direct Perception, each offering unique advantages and challenges in enhancing autonomous driving capabilities. However, there is a gap in research addressing integrating these approaches and understanding their relevance in diverse driving scenarios. This study introduces three distinct neural network models corresponding to Mediated Perception, Behavior Reflex, and Direct Perception approaches. We explore their significance across varying driving conditions, shedding light on the strengths and limitations of each approach. Our architecture fuses information from the base, future latent vector prediction, and auxiliary task networks, using global routing commands to select appropriate action sub-networks. We aim to provide insights into effectively utilizing diverse modeling strategies in autonomous driving by conducting experiments and evaluations. The results show that the ensemble model performs better than the individual approaches, suggesting that each modality contributes uniquely toward the performance of the overall model. Moreover, by exploring the significance of each modality, this study offers a roadmap for future research in autonomous driving, emphasizing the importance of leveraging multiple models to achieve robust performance.
Liked
zrz@andrew.cmu.edu
Beyond One Model Fits All: Ensemble Deep Learning for Autonomous Vehicles : Deep learning has revolutionized autonomous driving by enabling vehicles to perceive and interpret their surroundings with remarkable accuracy. This progress is attributed to various deep learning models, including Mediated Perception, Behavior Reflex, and Direct Perception, each offering unique advantages and challenges in enhancing autonomous driving capabilities. However, there is a gap in research addressing integrating these approaches and understanding their relevance in diverse driving scenarios. This study introduces three distinct neural network models corresponding to Mediated Perception, Behavior Reflex, and Direct Perception approaches. We explore their significance across varying driving conditions, shedding light on the strengths and limitations of each approach. Our architecture fuses information from the base, future latent vector prediction, and auxiliary task networks, using global routing commands to select appropriate action sub-networks. We aim to provide insights into effectively utilizing diverse modeling strategies in autonomous driving by conducting experiments and evaluations. The results show that the ensemble model performs better than the individual approaches, suggesting that each modality contributes uniquely toward the performance of the overall model. Moreover, by exploring the significance of each modality, this study offers a roadmap for future research in autonomous driving, emphasizing the importance of leveraging multiple models to achieve robust performance.
1
zrz@andrew.cmu.edu [SEP] Beyond One Model Fits All: Ensemble Deep Learning for Autonomous Vehicles : Deep learning has revolutionized autonomous driving by enabling vehicles to perceive and interpret their surroundings with remarkable accuracy. This progress is attributed to various deep learning models, including Mediated Perception, Behavior Reflex, and Direct Perception, each offering unique advantages and challenges in enhancing autonomous driving capabilities. However, there is a gap in research addressing integrating these approaches and understanding their relevance in diverse driving scenarios. This study introduces three distinct neural network models corresponding to Mediated Perception, Behavior Reflex, and Direct Perception approaches. We explore their significance across varying driving conditions, shedding light on the strengths and limitations of each approach. Our architecture fuses information from the base, future latent vector prediction, and auxiliary task networks, using global routing commands to select appropriate action sub-networks. We aim to provide insights into effectively utilizing diverse modeling strategies in autonomous driving by conducting experiments and evaluations. The results show that the ensemble model performs better than the individual approaches, suggesting that each modality contributes uniquely toward the performance of the overall model. Moreover, by exploring the significance of each modality, this study offers a roadmap for future research in autonomous driving, emphasizing the importance of leveraging multiple models to achieve robust performance.
303
Deep Learning and Its Applications to Machine Health Monitoring: A Survey
Since 2006, deep learning (DL) has become a rapidly growing research direction, redefining state-of-the-art performances in a wide range of areas such as object recognition, image segmentation, speech recognition and machine translation. In modern manufacturing systems, data-driven machine health monitoring is gaining in popularity due to the widespread deployment of low-cost sensors and their connection to the Internet. Meanwhile, deep learning provides useful tools for processing and analyzing these big machinery data. The main purpose of this paper is to review and summarize the emerging research work of deep learning on machine health monitoring. After the brief introduction of deep learning techniques, the applications of deep learning in machine health monitoring systems are reviewed mainly from the following aspects: Auto-encoder (AE) and its variants, Restricted Boltzmann Machines and its variants including Deep Belief Network (DBN) and Deep Boltzmann Machines (DBM), Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN). Finally, some new trends of DL-based machine health monitoring methods are discussed.
Liked
zrz@andrew.cmu.edu
Deep Learning and Its Applications to Machine Health Monitoring: A Survey : Since 2006, deep learning (DL) has become a rapidly growing research direction, redefining state-of-the-art performances in a wide range of areas such as object recognition, image segmentation, speech recognition and machine translation. In modern manufacturing systems, data-driven machine health monitoring is gaining in popularity due to the widespread deployment of low-cost sensors and their connection to the Internet. Meanwhile, deep learning provides useful tools for processing and analyzing these big machinery data. The main purpose of this paper is to review and summarize the emerging research work of deep learning on machine health monitoring. After the brief introduction of deep learning techniques, the applications of deep learning in machine health monitoring systems are reviewed mainly from the following aspects: Auto-encoder (AE) and its variants, Restricted Boltzmann Machines and its variants including Deep Belief Network (DBN) and Deep Boltzmann Machines (DBM), Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN). Finally, some new trends of DL-based machine health monitoring methods are discussed.
1
zrz@andrew.cmu.edu [SEP] Deep Learning and Its Applications to Machine Health Monitoring: A Survey : Since 2006, deep learning (DL) has become a rapidly growing research direction, redefining state-of-the-art performances in a wide range of areas such as object recognition, image segmentation, speech recognition and machine translation. In modern manufacturing systems, data-driven machine health monitoring is gaining in popularity due to the widespread deployment of low-cost sensors and their connection to the Internet. Meanwhile, deep learning provides useful tools for processing and analyzing these big machinery data. The main purpose of this paper is to review and summarize the emerging research work of deep learning on machine health monitoring. After the brief introduction of deep learning techniques, the applications of deep learning in machine health monitoring systems are reviewed mainly from the following aspects: Auto-encoder (AE) and its variants, Restricted Boltzmann Machines and its variants including Deep Belief Network (DBN) and Deep Boltzmann Machines (DBM), Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN). Finally, some new trends of DL-based machine health monitoring methods are discussed.
240
Can Machines Learn the True Probabilities?
When there exists uncertainty, AI machines are designed to make decisions so as to reach the best expected outcomes. Expectations are based on true facts about the objective environment the machines interact with, and those facts can be encoded into AI models in the form of true objective probability functions. Accordingly, AI models involve probabilistic machine learning in which the probabilities should be objectively interpreted. We prove under some basic assumptions when machines can learn the true objective probabilities, if any, and when machines cannot learn them.
Liked
zrz@andrew.cmu.edu
Can Machines Learn the True Probabilities? : When there exists uncertainty, AI machines are designed to make decisions so as to reach the best expected outcomes. Expectations are based on true facts about the objective environment the machines interact with, and those facts can be encoded into AI models in the form of true objective probability functions. Accordingly, AI models involve probabilistic machine learning in which the probabilities should be objectively interpreted. We prove under some basic assumptions when machines can learn the true objective probabilities, if any, and when machines cannot learn them.
1
zrz@andrew.cmu.edu [SEP] Can Machines Learn the True Probabilities? : When there exists uncertainty, AI machines are designed to make decisions so as to reach the best expected outcomes. Expectations are based on true facts about the objective environment the machines interact with, and those facts can be encoded into AI models in the form of true objective probability functions. Accordingly, AI models involve probabilistic machine learning in which the probabilities should be objectively interpreted. We prove under some basic assumptions when machines can learn the true objective probabilities, if any, and when machines cannot learn them.
66
Teaching Computer Vision for Ecology
Computer vision can accelerate ecology research by automating the analysis of raw imagery from sensors like camera traps, drones, and satellites. However, computer vision is an emerging discipline that is rarely taught to ecologists. This work discusses our experience teaching a diverse group of ecologists to prototype and evaluate computer vision systems in the context of an intensive hands-on summer workshop. We explain the workshop structure, discuss common challenges, and propose best practices. This document is intended for computer scientists who teach computer vision across disciplines, but it may also be useful to ecologists or other domain experts who are learning to use computer vision themselves.
Disliked
zrz@andrew.cmu.edu
Teaching Computer Vision for Ecology : Computer vision can accelerate ecology research by automating the analysis of raw imagery from sensors like camera traps, drones, and satellites. However, computer vision is an emerging discipline that is rarely taught to ecologists. This work discusses our experience teaching a diverse group of ecologists to prototype and evaluate computer vision systems in the context of an intensive hands-on summer workshop. We explain the workshop structure, discuss common challenges, and propose best practices. This document is intended for computer scientists who teach computer vision across disciplines, but it may also be useful to ecologists or other domain experts who are learning to use computer vision themselves.
0
zrz@andrew.cmu.edu [SEP] Teaching Computer Vision for Ecology : Computer vision can accelerate ecology research by automating the analysis of raw imagery from sensors like camera traps, drones, and satellites. However, computer vision is an emerging discipline that is rarely taught to ecologists. This work discusses our experience teaching a diverse group of ecologists to prototype and evaluate computer vision systems in the context of an intensive hands-on summer workshop. We explain the workshop structure, discuss common challenges, and propose best practices. This document is intended for computer scientists who teach computer vision across disciplines, but it may also be useful to ecologists or other domain experts who are learning to use computer vision themselves.
375
Semantic-Aware Ship Detection with Vision-Language Integration
Ship detection in remote sensing imagery is a critical task with wide-ranging applications, such as maritime activity monitoring, shipping logistics, and environmental studies. However, existing methods often struggle to capture fine-grained semantic information, limiting their effectiveness in complex scenarios. To address these challenges, we propose a novel detection framework that combines Vision-Language Models (VLMs) with a multi-scale adaptive sliding window strategy. To facilitate Semantic-Aware Ship Detection (SASD), we introduce ShipSem-VL, a specialized Vision-Language dataset designed to capture fine-grained ship attributes. We evaluate our framework through three well-defined tasks, providing a comprehensive analysis of its performance and demonstrating its effectiveness in advancing SASD from multiple perspectives.
Liked
zrz@andrew.cmu.edu
Semantic-Aware Ship Detection with Vision-Language Integration : Ship detection in remote sensing imagery is a critical task with wide-ranging applications, such as maritime activity monitoring, shipping logistics, and environmental studies. However, existing methods often struggle to capture fine-grained semantic information, limiting their effectiveness in complex scenarios. To address these challenges, we propose a novel detection framework that combines Vision-Language Models (VLMs) with a multi-scale adaptive sliding window strategy. To facilitate Semantic-Aware Ship Detection (SASD), we introduce ShipSem-VL, a specialized Vision-Language dataset designed to capture fine-grained ship attributes. We evaluate our framework through three well-defined tasks, providing a comprehensive analysis of its performance and demonstrating its effectiveness in advancing SASD from multiple perspectives.
1
zrz@andrew.cmu.edu [SEP] Semantic-Aware Ship Detection with Vision-Language Integration : Ship detection in remote sensing imagery is a critical task with wide-ranging applications, such as maritime activity monitoring, shipping logistics, and environmental studies. However, existing methods often struggle to capture fine-grained semantic information, limiting their effectiveness in complex scenarios. To address these challenges, we propose a novel detection framework that combines Vision-Language Models (VLMs) with a multi-scale adaptive sliding window strategy. To facilitate Semantic-Aware Ship Detection (SASD), we introduce ShipSem-VL, a specialized Vision-Language dataset designed to capture fine-grained ship attributes. We evaluate our framework through three well-defined tasks, providing a comprehensive analysis of its performance and demonstrating its effectiveness in advancing SASD from multiple perspectives.
347
When Machine Learning Meets Privacy: A Survey and Outlook
The newly emerged machine learning (e.g. deep learning) methods have become a strong driving force to revolutionize a wide range of industries, such as smart healthcare, financial technology, and surveillance systems. Meanwhile, privacy has emerged as a big concern in this machine learning-based artificial intelligence era. It is important to note that the problem of privacy preservation in the context of machine learning is quite different from that in traditional data privacy protection, as machine learning can act as both friend and foe. Currently, the work on the preservation of privacy and machine learning (ML) is still in an infancy stage, as most existing solutions only focus on privacy problems during the machine learning process. Therefore, a comprehensive study on the privacy preservation problems and machine learning is required. This paper surveys the state of the art in privacy issues and solutions for machine learning. The survey covers three categories of interactions between privacy and machine learning: (i) private machine learning, (ii) machine learning aided privacy protection, and (iii) machine learning-based privacy attack and corresponding protection schemes. The current research progress in each category is reviewed and the key challenges are identified. Finally, based on our in-depth analysis of the area of privacy and machine learning, we point out future research directions in this field.
Disliked
zrz@andrew.cmu.edu
When Machine Learning Meets Privacy: A Survey and Outlook : The newly emerged machine learning (e.g. deep learning) methods have become a strong driving force to revolutionize a wide range of industries, such as smart healthcare, financial technology, and surveillance systems. Meanwhile, privacy has emerged as a big concern in this machine learning-based artificial intelligence era. It is important to note that the problem of privacy preservation in the context of machine learning is quite different from that in traditional data privacy protection, as machine learning can act as both friend and foe. Currently, the work on the preservation of privacy and machine learning (ML) is still in an infancy stage, as most existing solutions only focus on privacy problems during the machine learning process. Therefore, a comprehensive study on the privacy preservation problems and machine learning is required. This paper surveys the state of the art in privacy issues and solutions for machine learning. The survey covers three categories of interactions between privacy and machine learning: (i) private machine learning, (ii) machine learning aided privacy protection, and (iii) machine learning-based privacy attack and corresponding protection schemes. The current research progress in each category is reviewed and the key challenges are identified. Finally, based on our in-depth analysis of the area of privacy and machine learning, we point out future research directions in this field.
0
zrz@andrew.cmu.edu [SEP] When Machine Learning Meets Privacy: A Survey and Outlook : The newly emerged machine learning (e.g. deep learning) methods have become a strong driving force to revolutionize a wide range of industries, such as smart healthcare, financial technology, and surveillance systems. Meanwhile, privacy has emerged as a big concern in this machine learning-based artificial intelligence era. It is important to note that the problem of privacy preservation in the context of machine learning is quite different from that in traditional data privacy protection, as machine learning can act as both friend and foe. Currently, the work on the preservation of privacy and machine learning (ML) is still in an infancy stage, as most existing solutions only focus on privacy problems during the machine learning process. Therefore, a comprehensive study on the privacy preservation problems and machine learning is required. This paper surveys the state of the art in privacy issues and solutions for machine learning. The survey covers three categories of interactions between privacy and machine learning: (i) private machine learning, (ii) machine learning aided privacy protection, and (iii) machine learning-based privacy attack and corresponding protection schemes. The current research progress in each category is reviewed and the key challenges are identified. Finally, based on our in-depth analysis of the area of privacy and machine learning, we point out future research directions in this field.
51
Machine Learning for Clinical Predictive Analytics
In this chapter, we provide a brief overview of applying machine learning techniques for clinical prediction tasks. We begin with a quick introduction to the concepts of machine learning and outline some of the most common machine learning algorithms. Next, we demonstrate how to apply the algorithms with appropriate toolkits to conduct machine learning experiments for clinical prediction tasks. The objectives of this chapter are to (1) understand the basics of machine learning techniques and the reasons behind why they are useful for solving clinical prediction problems, (2) understand the intuition behind some machine learning models, including regression, decision trees, and support vector machines, and (3) understand how to apply these models to clinical prediction problems using publicly available datasets via case studies.
Disliked
zrz@andrew.cmu.edu
Machine Learning for Clinical Predictive Analytics : In this chapter, we provide a brief overview of applying machine learning techniques for clinical prediction tasks. We begin with a quick introduction to the concepts of machine learning and outline some of the most common machine learning algorithms. Next, we demonstrate how to apply the algorithms with appropriate toolkits to conduct machine learning experiments for clinical prediction tasks. The objectives of this chapter are to (1) understand the basics of machine learning techniques and the reasons behind why they are useful for solving clinical prediction problems, (2) understand the intuition behind some machine learning models, including regression, decision trees, and support vector machines, and (3) understand how to apply these models to clinical prediction problems using publicly available datasets via case studies.
0
zrz@andrew.cmu.edu [SEP] Machine Learning for Clinical Predictive Analytics : In this chapter, we provide a brief overview of applying machine learning techniques for clinical prediction tasks. We begin with a quick introduction to the concepts of machine learning and outline some of the most common machine learning algorithms. Next, we demonstrate how to apply the algorithms with appropriate toolkits to conduct machine learning experiments for clinical prediction tasks. The objectives of this chapter are to (1) understand the basics of machine learning techniques and the reasons behind why they are useful for solving clinical prediction problems, (2) understand the intuition behind some machine learning models, including regression, decision trees, and support vector machines, and (3) understand how to apply these models to clinical prediction problems using publicly available datasets via case studies.
28
Enhancing camera surveillance using computer vision: a research note
$\mathbf{Purpose}$ - The growth of police operated surveillance cameras has out-paced the ability of humans to monitor them effectively. Computer vision is a possible solution. An ongoing research project on the application of computer vision within a municipal police department is described. The paper aims to discuss these issues. $\mathbf{Design/methodology/approach}$ - Following the demystification of computer vision technology, its potential for police agencies is developed within a focus on computer vision as a solution for two common surveillance camera tasks (live monitoring of multiple surveillance cameras and summarizing archived video files). Three unaddressed research questions (can specialized computer vision applications for law enforcement be developed at this time, how will computer vision be utilized within existing public safety camera monitoring rooms, and what are the system-wide impacts of a computer vision capability on local criminal justice systems) are considered. $\mathbf{Findings}$ - Despite computer vision becoming accessible to law enforcement agencies the impact of computer vision has not been discussed or adequately researched. There is little knowledge of computer vision or its potential in the field. $\mathbf{Originality/value}$ - This paper introduces and discusses computer vision from a law enforcement perspective and will be valuable to police personnel tasked with monitoring large camera networks and considering computer vision as a system upgrade.
Disliked
zrz@andrew.cmu.edu
Enhancing camera surveillance using computer vision: a research note : $\mathbf{Purpose}$ - The growth of police operated surveillance cameras has out-paced the ability of humans to monitor them effectively. Computer vision is a possible solution. An ongoing research project on the application of computer vision within a municipal police department is described. The paper aims to discuss these issues. $\mathbf{Design/methodology/approach}$ - Following the demystification of computer vision technology, its potential for police agencies is developed within a focus on computer vision as a solution for two common surveillance camera tasks (live monitoring of multiple surveillance cameras and summarizing archived video files). Three unaddressed research questions (can specialized computer vision applications for law enforcement be developed at this time, how will computer vision be utilized within existing public safety camera monitoring rooms, and what are the system-wide impacts of a computer vision capability on local criminal justice systems) are considered. $\mathbf{Findings}$ - Despite computer vision becoming accessible to law enforcement agencies the impact of computer vision has not been discussed or adequately researched. There is little knowledge of computer vision or its potential in the field. $\mathbf{Originality/value}$ - This paper introduces and discusses computer vision from a law enforcement perspective and will be valuable to police personnel tasked with monitoring large camera networks and considering computer vision as a system upgrade.
0
zrz@andrew.cmu.edu [SEP] Enhancing camera surveillance using computer vision: a research note : $\mathbf{Purpose}$ - The growth of police operated surveillance cameras has out-paced the ability of humans to monitor them effectively. Computer vision is a possible solution. An ongoing research project on the application of computer vision within a municipal police department is described. The paper aims to discuss these issues. $\mathbf{Design/methodology/approach}$ - Following the demystification of computer vision technology, its potential for police agencies is developed within a focus on computer vision as a solution for two common surveillance camera tasks (live monitoring of multiple surveillance cameras and summarizing archived video files). Three unaddressed research questions (can specialized computer vision applications for law enforcement be developed at this time, how will computer vision be utilized within existing public safety camera monitoring rooms, and what are the system-wide impacts of a computer vision capability on local criminal justice systems) are considered. $\mathbf{Findings}$ - Despite computer vision becoming accessible to law enforcement agencies the impact of computer vision has not been discussed or adequately researched. There is little knowledge of computer vision or its potential in the field. $\mathbf{Originality/value}$ - This paper introduces and discusses computer vision from a law enforcement perspective and will be valuable to police personnel tasked with monitoring large camera networks and considering computer vision as a system upgrade.
343
Meta-Learning: A Survey
Meta-learning, or learning to learn, is the science of systematically observing how different machine learning approaches perform on a wide range of learning tasks, and then learning from this experience, or meta-data, to learn new tasks much faster than otherwise possible. Not only does this dramatically speed up and improve the design of machine learning pipelines or neural architectures, it also allows us to replace hand-engineered algorithms with novel approaches learned in a data-driven way. In this chapter, we provide an overview of the state of the art in this fascinating and continuously evolving field.
Liked
zrz@andrew.cmu.edu
Meta-Learning: A Survey : Meta-learning, or learning to learn, is the science of systematically observing how different machine learning approaches perform on a wide range of learning tasks, and then learning from this experience, or meta-data, to learn new tasks much faster than otherwise possible. Not only does this dramatically speed up and improve the design of machine learning pipelines or neural architectures, it also allows us to replace hand-engineered algorithms with novel approaches learned in a data-driven way. In this chapter, we provide an overview of the state of the art in this fascinating and continuously evolving field.
1
zrz@andrew.cmu.edu [SEP] Meta-Learning: A Survey : Meta-learning, or learning to learn, is the science of systematically observing how different machine learning approaches perform on a wide range of learning tasks, and then learning from this experience, or meta-data, to learn new tasks much faster than otherwise possible. Not only does this dramatically speed up and improve the design of machine learning pipelines or neural architectures, it also allows us to replace hand-engineered algorithms with novel approaches learned in a data-driven way. In this chapter, we provide an overview of the state of the art in this fascinating and continuously evolving field.
87
Preparatory Manipulation Planning using Automatically Determined Single and Dual Arms
This paper presents a manipulation planning algorithm for robots to reorient objects. It automatically finds a sequence of robot motion that manipulates and prepares an object for specific tasks. Examples of the preparatory manipulation planning problems include reorienting an electric drill to cut holes, reorienting workpieces for assembly, and reorienting cargo for packing, etc. The proposed algorithm could plan single and dual arm manipulation sequences to solve the problems. The mechanism under the planner is a regrasp graph which encodes grasp configurations and object poses. The algorithms search the graph to find a sequence of robot motion to reorient objects. The planner is able to plan both single and dual arm manipulation. It could also automatically determine whether to use a single arm, dual arms, or their combinations to finish given tasks. The planner is examined by various humanoid robots like Nextage, HRP2Kai, HRP5P, etc., using both simulation and real-world experiments.
Liked
jechoi@andrew.cmu.edu
Preparatory Manipulation Planning using Automatically Determined Single and Dual Arms : This paper presents a manipulation planning algorithm for robots to reorient objects. It automatically finds a sequence of robot motion that manipulates and prepares an object for specific tasks. Examples of the preparatory manipulation planning problems include reorienting an electric drill to cut holes, reorienting workpieces for assembly, and reorienting cargo for packing, etc. The proposed algorithm could plan single and dual arm manipulation sequences to solve the problems. The mechanism under the planner is a regrasp graph which encodes grasp configurations and object poses. The algorithms search the graph to find a sequence of robot motion to reorient objects. The planner is able to plan both single and dual arm manipulation. It could also automatically determine whether to use a single arm, dual arms, or their combinations to finish given tasks. The planner is examined by various humanoid robots like Nextage, HRP2Kai, HRP5P, etc., using both simulation and real-world experiments.
1
jechoi@andrew.cmu.edu [SEP] Preparatory Manipulation Planning using Automatically Determined Single and Dual Arms : This paper presents a manipulation planning algorithm for robots to reorient objects. It automatically finds a sequence of robot motion that manipulates and prepares an object for specific tasks. Examples of the preparatory manipulation planning problems include reorienting an electric drill to cut holes, reorienting workpieces for assembly, and reorienting cargo for packing, etc. The proposed algorithm could plan single and dual arm manipulation sequences to solve the problems. The mechanism under the planner is a regrasp graph which encodes grasp configurations and object poses. The algorithms search the graph to find a sequence of robot motion to reorient objects. The planner is able to plan both single and dual arm manipulation. It could also automatically determine whether to use a single arm, dual arms, or their combinations to finish given tasks. The planner is examined by various humanoid robots like Nextage, HRP2Kai, HRP5P, etc., using both simulation and real-world experiments.
496
Arm Manipulation Planning of Tethered Tools with the Help of a Tool Balancer
Robotic manipulation of tethered tools is widely seen in robotic work cells. They may cause excess strain on the tool's cable or undesired entanglements with the robot's arms. This paper presents a manipulation planner with cable orientation constraints for tethered tools suspended by tool balancers. The planner uses orientation constraints to limit the bending of the balancer's cable while the robot manipulates a tool and places it in a desired pose. The constraints reduce entanglements and decrease the torque induced by the cable on the robot joints. Simulation and real-world experiments show that the constrained planner can successfully plan robot motions for the manipulation of suspended tethered tools preventing the robot from damaging the cable or getting its arms entangled, potentially avoiding accidents. The planner is expected to play promising roles in manufacturing cells.
Liked
jechoi@andrew.cmu.edu
Arm Manipulation Planning of Tethered Tools with the Help of a Tool Balancer : Robotic manipulation of tethered tools is widely seen in robotic work cells. They may cause excess strain on the tool's cable or undesired entanglements with the robot's arms. This paper presents a manipulation planner with cable orientation constraints for tethered tools suspended by tool balancers. The planner uses orientation constraints to limit the bending of the balancer's cable while the robot manipulates a tool and places it in a desired pose. The constraints reduce entanglements and decrease the torque induced by the cable on the robot joints. Simulation and real-world experiments show that the constrained planner can successfully plan robot motions for the manipulation of suspended tethered tools preventing the robot from damaging the cable or getting its arms entangled, potentially avoiding accidents. The planner is expected to play promising roles in manufacturing cells.
1
jechoi@andrew.cmu.edu [SEP] Arm Manipulation Planning of Tethered Tools with the Help of a Tool Balancer : Robotic manipulation of tethered tools is widely seen in robotic work cells. They may cause excess strain on the tool's cable or undesired entanglements with the robot's arms. This paper presents a manipulation planner with cable orientation constraints for tethered tools suspended by tool balancers. The planner uses orientation constraints to limit the bending of the balancer's cable while the robot manipulates a tool and places it in a desired pose. The constraints reduce entanglements and decrease the torque induced by the cable on the robot joints. Simulation and real-world experiments show that the constrained planner can successfully plan robot motions for the manipulation of suspended tethered tools preventing the robot from damaging the cable or getting its arms entangled, potentially avoiding accidents. The planner is expected to play promising roles in manufacturing cells.
529
Lidar for Autonomous Driving: The principles, challenges, and trends for automotive lidar and perception systems
Autonomous vehicles rely on their perception systems to acquire information about their immediate surroundings. It is necessary to detect the presence of other vehicles, pedestrians and other relevant entities. Safety concerns and the need for accurate estimations have led to the introduction of Light Detection and Ranging (LiDAR) systems in complement to the camera or radar-based perception systems. This article presents a review of state-of-the-art automotive LiDAR technologies and the perception algorithms used with those technologies. LiDAR systems are introduced first by analyzing the main components, from laser transmitter to its beam scanning mechanism. Advantages/disadvantages and the current status of various solutions are introduced and compared. Then, the specific perception pipeline for LiDAR data processing, from an autonomous vehicle perspective is detailed. The model-driven approaches and the emerging deep learning solutions are reviewed. Finally, we provide an overview of the limitations, challenges and trends for automotive LiDARs and perception systems.
Liked
zrz@andrew.cmu.edu
Lidar for Autonomous Driving: The principles, challenges, and trends for automotive lidar and perception systems : Autonomous vehicles rely on their perception systems to acquire information about their immediate surroundings. It is necessary to detect the presence of other vehicles, pedestrians and other relevant entities. Safety concerns and the need for accurate estimations have led to the introduction of Light Detection and Ranging (LiDAR) systems in complement to the camera or radar-based perception systems. This article presents a review of state-of-the-art automotive LiDAR technologies and the perception algorithms used with those technologies. LiDAR systems are introduced first by analyzing the main components, from laser transmitter to its beam scanning mechanism. Advantages/disadvantages and the current status of various solutions are introduced and compared. Then, the specific perception pipeline for LiDAR data processing, from an autonomous vehicle perspective is detailed. The model-driven approaches and the emerging deep learning solutions are reviewed. Finally, we provide an overview of the limitations, challenges and trends for automotive LiDARs and perception systems.
1
zrz@andrew.cmu.edu [SEP] Lidar for Autonomous Driving: The principles, challenges, and trends for automotive lidar and perception systems : Autonomous vehicles rely on their perception systems to acquire information about their immediate surroundings. It is necessary to detect the presence of other vehicles, pedestrians and other relevant entities. Safety concerns and the need for accurate estimations have led to the introduction of Light Detection and Ranging (LiDAR) systems in complement to the camera or radar-based perception systems. This article presents a review of state-of-the-art automotive LiDAR technologies and the perception algorithms used with those technologies. LiDAR systems are introduced first by analyzing the main components, from laser transmitter to its beam scanning mechanism. Advantages/disadvantages and the current status of various solutions are introduced and compared. Then, the specific perception pipeline for LiDAR data processing, from an autonomous vehicle perspective is detailed. The model-driven approaches and the emerging deep learning solutions are reviewed. Finally, we provide an overview of the limitations, challenges and trends for automotive LiDARs and perception systems.
296
Deep Gaussian Mixture Models
Deep learning is a hierarchical inference method formed by subsequent multiple layers of learning able to more efficiently describe complex relationships. In this work, Deep Gaussian Mixture Models are introduced and discussed. A Deep Gaussian Mixture model (DGMM) is a network of multiple layers of latent variables, where, at each layer, the variables follow a mixture of Gaussian distributions. Thus, the deep mixture model consists of a set of nested mixtures of linear models, which globally provide a nonlinear model able to describe the data in a very flexible way. In order to avoid overparameterized solutions, dimension reduction by factor models can be applied at each layer of the architecture thus resulting in deep mixtures of factor analysers.
Disliked
zrz@andrew.cmu.edu
Deep Gaussian Mixture Models : Deep learning is a hierarchical inference method formed by subsequent multiple layers of learning able to more efficiently describe complex relationships. In this work, Deep Gaussian Mixture Models are introduced and discussed. A Deep Gaussian Mixture model (DGMM) is a network of multiple layers of latent variables, where, at each layer, the variables follow a mixture of Gaussian distributions. Thus, the deep mixture model consists of a set of nested mixtures of linear models, which globally provide a nonlinear model able to describe the data in a very flexible way. In order to avoid overparameterized solutions, dimension reduction by factor models can be applied at each layer of the architecture thus resulting in deep mixtures of factor analysers.
0
zrz@andrew.cmu.edu [SEP] Deep Gaussian Mixture Models : Deep learning is a hierarchical inference method formed by subsequent multiple layers of learning able to more efficiently describe complex relationships. In this work, Deep Gaussian Mixture Models are introduced and discussed. A Deep Gaussian Mixture model (DGMM) is a network of multiple layers of latent variables, where, at each layer, the variables follow a mixture of Gaussian distributions. Thus, the deep mixture model consists of a set of nested mixtures of linear models, which globally provide a nonlinear model able to describe the data in a very flexible way. In order to avoid overparameterized solutions, dimension reduction by factor models can be applied at each layer of the architecture thus resulting in deep mixtures of factor analysers.
268
Experimental Characterization of Robot Arm Rigidity in Order to Be Used in Machining Operation
Attempts to install a rotating tool at the end of a robot arm poly-articulated date back twenty years, but these robots were not designed for that. Indeed, two essential features are necessary for machining: high rigidity and precision in a given workspace. The experimental results presented are the dynamic identification of a poly-articulated robot equipped with an integrated spindle. This study aims to highlight the influence of the geometric configuration of the robot arm on the overall stiffness of the system. The spindle is taken into account as an additional weight on board but also as a dynamical excitation for the robot KUKA KR_240_2. Study of the robotic machining vibrations shows the suitable directions of movement in milling process
Liked
jechoi@andrew.cmu.edu
Experimental Characterization of Robot Arm Rigidity in Order to Be Used in Machining Operation : Attempts to install a rotating tool at the end of a robot arm poly-articulated date back twenty years, but these robots were not designed for that. Indeed, two essential features are necessary for machining: high rigidity and precision in a given workspace. The experimental results presented are the dynamic identification of a poly-articulated robot equipped with an integrated spindle. This study aims to highlight the influence of the geometric configuration of the robot arm on the overall stiffness of the system. The spindle is taken into account as an additional weight on board but also as a dynamical excitation for the robot KUKA KR_240_2. Study of the robotic machining vibrations shows the suitable directions of movement in milling process
1
jechoi@andrew.cmu.edu [SEP] Experimental Characterization of Robot Arm Rigidity in Order to Be Used in Machining Operation : Attempts to install a rotating tool at the end of a robot arm poly-articulated date back twenty years, but these robots were not designed for that. Indeed, two essential features are necessary for machining: high rigidity and precision in a given workspace. The experimental results presented are the dynamic identification of a poly-articulated robot equipped with an integrated spindle. This study aims to highlight the influence of the geometric configuration of the robot arm on the overall stiffness of the system. The spindle is taken into account as an additional weight on board but also as a dynamical excitation for the robot KUKA KR_240_2. Study of the robotic machining vibrations shows the suitable directions of movement in milling process
526
WiCV 2020: The Seventh Women In Computer Vision Workshop
In this paper we present the details of Women in Computer Vision Workshop - WiCV 2020, organized in alongside virtual CVPR 2020. This event aims at encouraging the women researchers in the field of computer vision. It provides a voice to a minority (female) group in computer vision community and focuses on increasingly the visibility of these researchers, both in academia and industry. WiCV believes that such an event can play an important role in lowering the gender imbalance in the field of computer vision. WiCV is organized each year where it provides a.) opportunity for collaboration with between researchers b.) mentorship to female junior researchers c.) financial support to presenters to overcome monetary burden and d.) large and diverse choice of role models, who can serve as examples to younger researchers at the beginning of their careers. In this paper, we present a report on the workshop program, trends over the past years, a summary of statistics regarding presenters, attendees, and sponsorship for the current workshop.
Disliked
zrz@andrew.cmu.edu
WiCV 2020: The Seventh Women In Computer Vision Workshop : In this paper we present the details of Women in Computer Vision Workshop - WiCV 2020, organized in alongside virtual CVPR 2020. This event aims at encouraging the women researchers in the field of computer vision. It provides a voice to a minority (female) group in computer vision community and focuses on increasingly the visibility of these researchers, both in academia and industry. WiCV believes that such an event can play an important role in lowering the gender imbalance in the field of computer vision. WiCV is organized each year where it provides a.) opportunity for collaboration with between researchers b.) mentorship to female junior researchers c.) financial support to presenters to overcome monetary burden and d.) large and diverse choice of role models, who can serve as examples to younger researchers at the beginning of their careers. In this paper, we present a report on the workshop program, trends over the past years, a summary of statistics regarding presenters, attendees, and sponsorship for the current workshop.
0
zrz@andrew.cmu.edu [SEP] WiCV 2020: The Seventh Women In Computer Vision Workshop : In this paper we present the details of Women in Computer Vision Workshop - WiCV 2020, organized in alongside virtual CVPR 2020. This event aims at encouraging the women researchers in the field of computer vision. It provides a voice to a minority (female) group in computer vision community and focuses on increasingly the visibility of these researchers, both in academia and industry. WiCV believes that such an event can play an important role in lowering the gender imbalance in the field of computer vision. WiCV is organized each year where it provides a.) opportunity for collaboration with between researchers b.) mentorship to female junior researchers c.) financial support to presenters to overcome monetary burden and d.) large and diverse choice of role models, who can serve as examples to younger researchers at the beginning of their careers. In this paper, we present a report on the workshop program, trends over the past years, a summary of statistics regarding presenters, attendees, and sponsorship for the current workshop.
373
Kinematic Optimization of a Robotic Arm for Automation Tasks with Human Demonstration
Robotic arms are highly common in various automation processes such as manufacturing lines. However, these highly capable robots are usually degraded to simple repetitive tasks such as pick-and-place. On the other hand, designing an optimal robot for one specific task consumes large resources of engineering time and costs. In this paper, we propose a novel concept for optimizing the fitness of a robotic arm to perform a specific task based on human demonstration. Fitness of a robot arm is a measure of its ability to follow recorded human arm and hand paths. The optimization is conducted using a modified variant of the Particle Swarm Optimization for the robot design problem. In the proposed approach, we generate an optimal robot design along with the required path to complete the task. The approach could reduce the time-to-market of robotic arms and enable the standardization of modular robotic parts. Novice users could easily apply a minimal robot arm to various tasks. Two test cases of common manufacturing tasks are presented yielding optimal designs and reduced computational effort by up to 92%.
Liked
jechoi@andrew.cmu.edu
Kinematic Optimization of a Robotic Arm for Automation Tasks with Human Demonstration : Robotic arms are highly common in various automation processes such as manufacturing lines. However, these highly capable robots are usually degraded to simple repetitive tasks such as pick-and-place. On the other hand, designing an optimal robot for one specific task consumes large resources of engineering time and costs. In this paper, we propose a novel concept for optimizing the fitness of a robotic arm to perform a specific task based on human demonstration. Fitness of a robot arm is a measure of its ability to follow recorded human arm and hand paths. The optimization is conducted using a modified variant of the Particle Swarm Optimization for the robot design problem. In the proposed approach, we generate an optimal robot design along with the required path to complete the task. The approach could reduce the time-to-market of robotic arms and enable the standardization of modular robotic parts. Novice users could easily apply a minimal robot arm to various tasks. Two test cases of common manufacturing tasks are presented yielding optimal designs and reduced computational effort by up to 92%.
1
jechoi@andrew.cmu.edu [SEP] Kinematic Optimization of a Robotic Arm for Automation Tasks with Human Demonstration : Robotic arms are highly common in various automation processes such as manufacturing lines. However, these highly capable robots are usually degraded to simple repetitive tasks such as pick-and-place. On the other hand, designing an optimal robot for one specific task consumes large resources of engineering time and costs. In this paper, we propose a novel concept for optimizing the fitness of a robotic arm to perform a specific task based on human demonstration. Fitness of a robot arm is a measure of its ability to follow recorded human arm and hand paths. The optimization is conducted using a modified variant of the Particle Swarm Optimization for the robot design problem. In the proposed approach, we generate an optimal robot design along with the required path to complete the task. The approach could reduce the time-to-market of robotic arms and enable the standardization of modular robotic parts. Novice users could easily apply a minimal robot arm to various tasks. Two test cases of common manufacturing tasks are presented yielding optimal designs and reduced computational effort by up to 92%.
390
Public Perceptions of Autonomous Vehicles: A Survey of Pedestrians and Cyclists in Pittsburgh
This study investigates how autonomous vehicle(AV) technology is perceived by pedestrians and bicyclists in Pittsburgh. Using survey data from over 1200 respondents, the research explores the interplay between demographics, AV interactions, infrastructural readiness, safety perceptions, and trust. Findings highlight demographic divides, infrastructure gaps, and the crucial role of communication and education in AV adoption.
Liked
zrz@andrew.cmu.edu
Public Perceptions of Autonomous Vehicles: A Survey of Pedestrians and Cyclists in Pittsburgh : This study investigates how autonomous vehicle(AV) technology is perceived by pedestrians and bicyclists in Pittsburgh. Using survey data from over 1200 respondents, the research explores the interplay between demographics, AV interactions, infrastructural readiness, safety perceptions, and trust. Findings highlight demographic divides, infrastructure gaps, and the crucial role of communication and education in AV adoption.
1
zrz@andrew.cmu.edu [SEP] Public Perceptions of Autonomous Vehicles: A Survey of Pedestrians and Cyclists in Pittsburgh : This study investigates how autonomous vehicle(AV) technology is perceived by pedestrians and bicyclists in Pittsburgh. Using survey data from over 1200 respondents, the research explores the interplay between demographics, AV interactions, infrastructural readiness, safety perceptions, and trust. Findings highlight demographic divides, infrastructure gaps, and the crucial role of communication and education in AV adoption.
301
Reflective VLM Planning for Dual-Arm Desktop Cleaning: Bridging Open-Vocabulary Perception and Precise Manipulation
Desktop cleaning demands open-vocabulary recognition and precise manipulation for heterogeneous debris. We propose a hierarchical framework integrating reflective Vision-Language Model (VLM) planning with dual-arm execution via structured scene representation. Grounded-SAM2 facilitates open-vocabulary detection, while a memory-augmented VLM generates, critiques, and revises manipulation sequences. These sequences are converted into parametric trajectories for five primitives executed by coordinated Franka arms. Evaluated in simulated scenarios, our system achieving 87.2% task completion, a 28.8% improvement over static VLM and 36.2% over single-arm baselines. Structured memory integration proves crucial for robust, generalizable manipulation while maintaining real-time control performance.
Liked
jechoi@andrew.cmu.edu
Reflective VLM Planning for Dual-Arm Desktop Cleaning: Bridging Open-Vocabulary Perception and Precise Manipulation : Desktop cleaning demands open-vocabulary recognition and precise manipulation for heterogeneous debris. We propose a hierarchical framework integrating reflective Vision-Language Model (VLM) planning with dual-arm execution via structured scene representation. Grounded-SAM2 facilitates open-vocabulary detection, while a memory-augmented VLM generates, critiques, and revises manipulation sequences. These sequences are converted into parametric trajectories for five primitives executed by coordinated Franka arms. Evaluated in simulated scenarios, our system achieving 87.2% task completion, a 28.8% improvement over static VLM and 36.2% over single-arm baselines. Structured memory integration proves crucial for robust, generalizable manipulation while maintaining real-time control performance.
1
jechoi@andrew.cmu.edu [SEP] Reflective VLM Planning for Dual-Arm Desktop Cleaning: Bridging Open-Vocabulary Perception and Precise Manipulation : Desktop cleaning demands open-vocabulary recognition and precise manipulation for heterogeneous debris. We propose a hierarchical framework integrating reflective Vision-Language Model (VLM) planning with dual-arm execution via structured scene representation. Grounded-SAM2 facilitates open-vocabulary detection, while a memory-augmented VLM generates, critiques, and revises manipulation sequences. These sequences are converted into parametric trajectories for five primitives executed by coordinated Franka arms. Evaluated in simulated scenarios, our system achieving 87.2% task completion, a 28.8% improvement over static VLM and 36.2% over single-arm baselines. Structured memory integration proves crucial for robust, generalizable manipulation while maintaining real-time control performance.
556
DAG-Plan: Generating Directed Acyclic Dependency Graphs for Dual-Arm Cooperative Planning
Dual-arm robots offer enhanced versatility and efficiency over single-arm counterparts by enabling concurrent manipulation of multiple objects or cooperative execution of tasks using both arms. However, the coordination of dual-arm systems for long-horizon tasks continues to pose significant challenges, stemming from the intricate temporal and spatial dependencies among sub-tasks, necessitating intelligent decisions regarding the allocation of actions between arms and their optimal execution order. Existing task planning methods predominantly focus on single-arm robots or rely on predefined bimanual operations to use large language models (LLMs) generate task sequence with linear temporal dependency, failing to fully leverage the capabilities of dual-arm systems. To address this limitation, we introduce DAG-Plan, a structured task planning framework tailored for dual-arm robots. DAG-Plan harnesses LLMs to decompose intricate tasks into actionable sub-tasks represented as nodes within a directed acyclic graph (DAG). Critically, DAG-Plan dynamically assigns these sub-tasks to the appropriate arm based on real-time environmental observations, enabling parallel and adaptive execution. We evaluate DAG-Plan on the Dual-Arm Kitchen Benchmark, comprising 5 sequential tasks with 44 sub-tasks. Extensive experiments demonstrate the superiority of DAG-Plan over directly using LLM to generate linear task sequence, achieving 52.8% higher efficiency compared to the single-arm task planning and 48% higher success rate of the dual-arm task planning. Compared to iterative methods, DAG-Plan improving execution efficiency 84.1% due to its fewer query time. More demos and information are available on https://sites.google.com/view/dag-plan.
Liked
jechoi@andrew.cmu.edu
DAG-Plan: Generating Directed Acyclic Dependency Graphs for Dual-Arm Cooperative Planning : Dual-arm robots offer enhanced versatility and efficiency over single-arm counterparts by enabling concurrent manipulation of multiple objects or cooperative execution of tasks using both arms. However, the coordination of dual-arm systems for long-horizon tasks continues to pose significant challenges, stemming from the intricate temporal and spatial dependencies among sub-tasks, necessitating intelligent decisions regarding the allocation of actions between arms and their optimal execution order. Existing task planning methods predominantly focus on single-arm robots or rely on predefined bimanual operations to use large language models (LLMs) generate task sequence with linear temporal dependency, failing to fully leverage the capabilities of dual-arm systems. To address this limitation, we introduce DAG-Plan, a structured task planning framework tailored for dual-arm robots. DAG-Plan harnesses LLMs to decompose intricate tasks into actionable sub-tasks represented as nodes within a directed acyclic graph (DAG). Critically, DAG-Plan dynamically assigns these sub-tasks to the appropriate arm based on real-time environmental observations, enabling parallel and adaptive execution. We evaluate DAG-Plan on the Dual-Arm Kitchen Benchmark, comprising 5 sequential tasks with 44 sub-tasks. Extensive experiments demonstrate the superiority of DAG-Plan over directly using LLM to generate linear task sequence, achieving 52.8% higher efficiency compared to the single-arm task planning and 48% higher success rate of the dual-arm task planning. Compared to iterative methods, DAG-Plan improving execution efficiency 84.1% due to its fewer query time. More demos and information are available on https://sites.google.com/view/dag-plan.
1
jechoi@andrew.cmu.edu [SEP] DAG-Plan: Generating Directed Acyclic Dependency Graphs for Dual-Arm Cooperative Planning : Dual-arm robots offer enhanced versatility and efficiency over single-arm counterparts by enabling concurrent manipulation of multiple objects or cooperative execution of tasks using both arms. However, the coordination of dual-arm systems for long-horizon tasks continues to pose significant challenges, stemming from the intricate temporal and spatial dependencies among sub-tasks, necessitating intelligent decisions regarding the allocation of actions between arms and their optimal execution order. Existing task planning methods predominantly focus on single-arm robots or rely on predefined bimanual operations to use large language models (LLMs) generate task sequence with linear temporal dependency, failing to fully leverage the capabilities of dual-arm systems. To address this limitation, we introduce DAG-Plan, a structured task planning framework tailored for dual-arm robots. DAG-Plan harnesses LLMs to decompose intricate tasks into actionable sub-tasks represented as nodes within a directed acyclic graph (DAG). Critically, DAG-Plan dynamically assigns these sub-tasks to the appropriate arm based on real-time environmental observations, enabling parallel and adaptive execution. We evaluate DAG-Plan on the Dual-Arm Kitchen Benchmark, comprising 5 sequential tasks with 44 sub-tasks. Extensive experiments demonstrate the superiority of DAG-Plan over directly using LLM to generate linear task sequence, achieving 52.8% higher efficiency compared to the single-arm task planning and 48% higher success rate of the dual-arm task planning. Compared to iterative methods, DAG-Plan improving execution efficiency 84.1% due to its fewer query time. More demos and information are available on https://sites.google.com/view/dag-plan.
487
Pen and Paper Exercises in Machine Learning
This is a collection of (mostly) pen-and-paper exercises in machine learning. The exercises are on the following topics: linear algebra, optimisation, directed graphical models, undirected graphical models, expressive power of graphical models, factor graphs and message passing, inference for hidden Markov models, model-based learning (including ICA and unnormalised models), sampling and Monte-Carlo integration, and variational inference.
Disliked
zrz@andrew.cmu.edu
Pen and Paper Exercises in Machine Learning : This is a collection of (mostly) pen-and-paper exercises in machine learning. The exercises are on the following topics: linear algebra, optimisation, directed graphical models, undirected graphical models, expressive power of graphical models, factor graphs and message passing, inference for hidden Markov models, model-based learning (including ICA and unnormalised models), sampling and Monte-Carlo integration, and variational inference.
0
zrz@andrew.cmu.edu [SEP] Pen and Paper Exercises in Machine Learning : This is a collection of (mostly) pen-and-paper exercises in machine learning. The exercises are on the following topics: linear algebra, optimisation, directed graphical models, undirected graphical models, expressive power of graphical models, factor graphs and message passing, inference for hidden Markov models, model-based learning (including ICA and unnormalised models), sampling and Monte-Carlo integration, and variational inference.
122
How to deal with glare for improved perception of Autonomous Vehicles
Vision sensors are versatile and can capture a wide range of visual cues, such as color, texture, shape, and depth. This versatility, along with the relatively inexpensive availability of machine vision cameras, played an important role in adopting vision-based environment perception systems in autonomous vehicles (AVs). However, vision-based perception systems can be easily affected by glare in the presence of a bright source of light, such as the sun or the headlights of the oncoming vehicle at night or simply by light reflecting off snow or ice-covered surfaces; scenarios encountered frequently during driving. In this paper, we investigate various glare reduction techniques, including the proposed saturated pixel-aware glare reduction technique for improved performance of the computer vision (CV) tasks employed by the perception layer of AVs. We evaluate these glare reduction methods based on various performance metrics of the CV algorithms used by the perception layer. Specifically, we considered object detection, object recognition, object tracking, depth estimation, and lane detection which are crucial for autonomous driving. The experimental findings validate the efficacy of the proposed glare reduction approach, showcasing enhanced performance across diverse perception tasks and remarkable resilience against varying levels of glare.
Liked
zrz@andrew.cmu.edu
How to deal with glare for improved perception of Autonomous Vehicles : Vision sensors are versatile and can capture a wide range of visual cues, such as color, texture, shape, and depth. This versatility, along with the relatively inexpensive availability of machine vision cameras, played an important role in adopting vision-based environment perception systems in autonomous vehicles (AVs). However, vision-based perception systems can be easily affected by glare in the presence of a bright source of light, such as the sun or the headlights of the oncoming vehicle at night or simply by light reflecting off snow or ice-covered surfaces; scenarios encountered frequently during driving. In this paper, we investigate various glare reduction techniques, including the proposed saturated pixel-aware glare reduction technique for improved performance of the computer vision (CV) tasks employed by the perception layer of AVs. We evaluate these glare reduction methods based on various performance metrics of the CV algorithms used by the perception layer. Specifically, we considered object detection, object recognition, object tracking, depth estimation, and lane detection which are crucial for autonomous driving. The experimental findings validate the efficacy of the proposed glare reduction approach, showcasing enhanced performance across diverse perception tasks and remarkable resilience against varying levels of glare.
1
zrz@andrew.cmu.edu [SEP] How to deal with glare for improved perception of Autonomous Vehicles : Vision sensors are versatile and can capture a wide range of visual cues, such as color, texture, shape, and depth. This versatility, along with the relatively inexpensive availability of machine vision cameras, played an important role in adopting vision-based environment perception systems in autonomous vehicles (AVs). However, vision-based perception systems can be easily affected by glare in the presence of a bright source of light, such as the sun or the headlights of the oncoming vehicle at night or simply by light reflecting off snow or ice-covered surfaces; scenarios encountered frequently during driving. In this paper, we investigate various glare reduction techniques, including the proposed saturated pixel-aware glare reduction technique for improved performance of the computer vision (CV) tasks employed by the perception layer of AVs. We evaluate these glare reduction methods based on various performance metrics of the CV algorithms used by the perception layer. Specifically, we considered object detection, object recognition, object tracking, depth estimation, and lane detection which are crucial for autonomous driving. The experimental findings validate the efficacy of the proposed glare reduction approach, showcasing enhanced performance across diverse perception tasks and remarkable resilience against varying levels of glare.
300
Multiband NFC for High-Throughput Wireless Computer Vision Sensor Network
Vision sensors lie in the heart of computer vision. In many computer vision applications, such as AR/VR, non-contacting near-field communication (NFC) with high throughput is required to transfer information to algorithms. In this work, we proposed a novel NFC system which utilizes multiple frequency bands to achieve high throughput.
Liked
zrz@andrew.cmu.edu
Multiband NFC for High-Throughput Wireless Computer Vision Sensor Network : Vision sensors lie in the heart of computer vision. In many computer vision applications, such as AR/VR, non-contacting near-field communication (NFC) with high throughput is required to transfer information to algorithms. In this work, we proposed a novel NFC system which utilizes multiple frequency bands to achieve high throughput.
1
zrz@andrew.cmu.edu [SEP] Multiband NFC for High-Throughput Wireless Computer Vision Sensor Network : Vision sensors lie in the heart of computer vision. In many computer vision applications, such as AR/VR, non-contacting near-field communication (NFC) with high throughput is required to transfer information to algorithms. In this work, we proposed a novel NFC system which utilizes multiple frequency bands to achieve high throughput.
337
PACC: A Passive-Arm Approach for High-Payload Collaborative Carrying with Quadruped Robots Using Model Predictive Control
In this paper, we introduce the concept of using passive arm structures with intrinsic impedance for robot-robot and human-robot collaborative carrying with quadruped robots. The concept is meant for a leader-follower task and takes a minimalist approach that focuses on exploiting the robots' payload capabilities and reducing energy consumption, without compromising the robot locomotion capabilities. We introduce a preliminary arm mechanical design and describe how to use its joint displacements to guide the robot's motion. To control the robot's locomotion, we propose a decentralized Model Predictive Controller that incorporates an approximation of the arm dynamics and the estimation of the external forces from the collaborative carrying. We validate the overall system experimentally by performing both robot-robot and human-robot collaborative carrying on a stair-like obstacle and on rough terrain.
Disliked
jechoi@andrew.cmu.edu
PACC: A Passive-Arm Approach for High-Payload Collaborative Carrying with Quadruped Robots Using Model Predictive Control : In this paper, we introduce the concept of using passive arm structures with intrinsic impedance for robot-robot and human-robot collaborative carrying with quadruped robots. The concept is meant for a leader-follower task and takes a minimalist approach that focuses on exploiting the robots' payload capabilities and reducing energy consumption, without compromising the robot locomotion capabilities. We introduce a preliminary arm mechanical design and describe how to use its joint displacements to guide the robot's motion. To control the robot's locomotion, we propose a decentralized Model Predictive Controller that incorporates an approximation of the arm dynamics and the estimation of the external forces from the collaborative carrying. We validate the overall system experimentally by performing both robot-robot and human-robot collaborative carrying on a stair-like obstacle and on rough terrain.
0
jechoi@andrew.cmu.edu [SEP] PACC: A Passive-Arm Approach for High-Payload Collaborative Carrying with Quadruped Robots Using Model Predictive Control : In this paper, we introduce the concept of using passive arm structures with intrinsic impedance for robot-robot and human-robot collaborative carrying with quadruped robots. The concept is meant for a leader-follower task and takes a minimalist approach that focuses on exploiting the robots' payload capabilities and reducing energy consumption, without compromising the robot locomotion capabilities. We introduce a preliminary arm mechanical design and describe how to use its joint displacements to guide the robot's motion. To control the robot's locomotion, we propose a decentralized Model Predictive Controller that incorporates an approximation of the arm dynamics and the estimation of the external forces from the collaborative carrying. We validate the overall system experimentally by performing both robot-robot and human-robot collaborative carrying on a stair-like obstacle and on rough terrain.
23
Tuning Learning Rates with the Cumulative-Learning Constant
This paper introduces a novel method for optimizing learning rates in machine learning. A previously unrecognized proportionality between learning rates and dataset sizes is discovered, providing valuable insights into how dataset scale influences training dynamics. Additionally, a cumulative learning constant is identified, offering a framework for designing and optimizing advanced learning rate schedules. These findings have the potential to enhance training efficiency and performance across a wide range of machine learning applications.
Liked
zrz@andrew.cmu.edu
Tuning Learning Rates with the Cumulative-Learning Constant : This paper introduces a novel method for optimizing learning rates in machine learning. A previously unrecognized proportionality between learning rates and dataset sizes is discovered, providing valuable insights into how dataset scale influences training dynamics. Additionally, a cumulative learning constant is identified, offering a framework for designing and optimizing advanced learning rate schedules. These findings have the potential to enhance training efficiency and performance across a wide range of machine learning applications.
1
zrz@andrew.cmu.edu [SEP] Tuning Learning Rates with the Cumulative-Learning Constant : This paper introduces a novel method for optimizing learning rates in machine learning. A previously unrecognized proportionality between learning rates and dataset sizes is discovered, providing valuable insights into how dataset scale influences training dynamics. Additionally, a cumulative learning constant is identified, offering a framework for designing and optimizing advanced learning rate schedules. These findings have the potential to enhance training efficiency and performance across a wide range of machine learning applications.
43
Multi-agent Collaborative Perception for Robotic Fleet: A Systematic Review
Collaborative perception in multi-robot fleets is a way to incorporate the power of unity in robotic fleets. Collaborative perception refers to the collective ability of multiple entities or agents to share and integrate their sensory information for a more comprehensive understanding of their environment. In other words, it involves the collaboration and fusion of data from various sensors or sources to enhance perception and decision-making capabilities. By combining data from diverse sources, such as cameras, lidar, radar, or other sensors, the system can create a more accurate and robust representation of the environment. In this review paper, we have summarized findings from 20+ research papers on collaborative perception. Moreover, we discuss testing and evaluation frameworks commonly accepted in academia and industry for autonomous vehicles and autonomous mobile robots. Our experiments with the trivial perception module show an improvement of over 200% with collaborative perception compared to individual robot perception. Here's our GitHub repository that shows the benefits of collaborative perception: https://github.com/synapsemobility/synapseBEV
Liked
zrz@andrew.cmu.edu
Multi-agent Collaborative Perception for Robotic Fleet: A Systematic Review : Collaborative perception in multi-robot fleets is a way to incorporate the power of unity in robotic fleets. Collaborative perception refers to the collective ability of multiple entities or agents to share and integrate their sensory information for a more comprehensive understanding of their environment. In other words, it involves the collaboration and fusion of data from various sensors or sources to enhance perception and decision-making capabilities. By combining data from diverse sources, such as cameras, lidar, radar, or other sensors, the system can create a more accurate and robust representation of the environment. In this review paper, we have summarized findings from 20+ research papers on collaborative perception. Moreover, we discuss testing and evaluation frameworks commonly accepted in academia and industry for autonomous vehicles and autonomous mobile robots. Our experiments with the trivial perception module show an improvement of over 200% with collaborative perception compared to individual robot perception. Here's our GitHub repository that shows the benefits of collaborative perception: https://github.com/synapsemobility/synapseBEV
1
zrz@andrew.cmu.edu [SEP] Multi-agent Collaborative Perception for Robotic Fleet: A Systematic Review : Collaborative perception in multi-robot fleets is a way to incorporate the power of unity in robotic fleets. Collaborative perception refers to the collective ability of multiple entities or agents to share and integrate their sensory information for a more comprehensive understanding of their environment. In other words, it involves the collaboration and fusion of data from various sensors or sources to enhance perception and decision-making capabilities. By combining data from diverse sources, such as cameras, lidar, radar, or other sensors, the system can create a more accurate and robust representation of the environment. In this review paper, we have summarized findings from 20+ research papers on collaborative perception. Moreover, we discuss testing and evaluation frameworks commonly accepted in academia and industry for autonomous vehicles and autonomous mobile robots. Our experiments with the trivial perception module show an improvement of over 200% with collaborative perception compared to individual robot perception. Here's our GitHub repository that shows the benefits of collaborative perception: https://github.com/synapsemobility/synapseBEV
279
V-MAO: Generative Modeling for Multi-Arm Manipulation of Articulated Objects
Manipulating articulated objects requires multiple robot arms in general. It is challenging to enable multiple robot arms to collaboratively complete manipulation tasks on articulated objects. In this paper, we present $\textbf{V-MAO}$, a framework for learning multi-arm manipulation of articulated objects. Our framework includes a variational generative model that learns contact point distribution over object rigid parts for each robot arm. The training signal is obtained from interaction with the simulation environment which is enabled by planning and a novel formulation of object-centric control for articulated objects. We deploy our framework in a customized MuJoCo simulation environment and demonstrate that our framework achieves a high success rate on six different objects and two different robots. We also show that generative modeling can effectively learn the contact point distribution on articulated objects.
Liked
jechoi@andrew.cmu.edu
V-MAO: Generative Modeling for Multi-Arm Manipulation of Articulated Objects : Manipulating articulated objects requires multiple robot arms in general. It is challenging to enable multiple robot arms to collaboratively complete manipulation tasks on articulated objects. In this paper, we present $\textbf{V-MAO}$, a framework for learning multi-arm manipulation of articulated objects. Our framework includes a variational generative model that learns contact point distribution over object rigid parts for each robot arm. The training signal is obtained from interaction with the simulation environment which is enabled by planning and a novel formulation of object-centric control for articulated objects. We deploy our framework in a customized MuJoCo simulation environment and demonstrate that our framework achieves a high success rate on six different objects and two different robots. We also show that generative modeling can effectively learn the contact point distribution on articulated objects.
1
jechoi@andrew.cmu.edu [SEP] V-MAO: Generative Modeling for Multi-Arm Manipulation of Articulated Objects : Manipulating articulated objects requires multiple robot arms in general. It is challenging to enable multiple robot arms to collaboratively complete manipulation tasks on articulated objects. In this paper, we present $\textbf{V-MAO}$, a framework for learning multi-arm manipulation of articulated objects. Our framework includes a variational generative model that learns contact point distribution over object rigid parts for each robot arm. The training signal is obtained from interaction with the simulation environment which is enabled by planning and a novel formulation of object-centric control for articulated objects. We deploy our framework in a customized MuJoCo simulation environment and demonstrate that our framework achieves a high success rate on six different objects and two different robots. We also show that generative modeling can effectively learn the contact point distribution on articulated objects.
498
Deep Embedding Kernel
In this paper, we propose a novel supervised learning method that is called Deep Embedding Kernel (DEK). DEK combines the advantages of deep learning and kernel methods in a unified framework. More specifically, DEK is a learnable kernel represented by a newly designed deep architecture. Compared with pre-defined kernels, this kernel can be explicitly trained to map data to an optimized high-level feature space where data may have favorable features toward the application. Compared with typical deep learning using SoftMax or logistic regression as the top layer, DEK is expected to be more generalizable to new data. Experimental results show that DEK has superior performance than typical machine learning methods in identity detection, classification, regression, dimension reduction, and transfer learning.
Disliked
zrz@andrew.cmu.edu
Deep Embedding Kernel : In this paper, we propose a novel supervised learning method that is called Deep Embedding Kernel (DEK). DEK combines the advantages of deep learning and kernel methods in a unified framework. More specifically, DEK is a learnable kernel represented by a newly designed deep architecture. Compared with pre-defined kernels, this kernel can be explicitly trained to map data to an optimized high-level feature space where data may have favorable features toward the application. Compared with typical deep learning using SoftMax or logistic regression as the top layer, DEK is expected to be more generalizable to new data. Experimental results show that DEK has superior performance than typical machine learning methods in identity detection, classification, regression, dimension reduction, and transfer learning.
0
zrz@andrew.cmu.edu [SEP] Deep Embedding Kernel : In this paper, we propose a novel supervised learning method that is called Deep Embedding Kernel (DEK). DEK combines the advantages of deep learning and kernel methods in a unified framework. More specifically, DEK is a learnable kernel represented by a newly designed deep architecture. Compared with pre-defined kernels, this kernel can be explicitly trained to map data to an optimized high-level feature space where data may have favorable features toward the application. Compared with typical deep learning using SoftMax or logistic regression as the top layer, DEK is expected to be more generalizable to new data. Experimental results show that DEK has superior performance than typical machine learning methods in identity detection, classification, regression, dimension reduction, and transfer learning.
215
Development of an Intuitive Foot-Machine Interface for Robotic Surgery
The human-machine interface is of critical importance for master-slave control of the robotic system for surgery, in which current systems offer the control or two robotic arms teleoperated by the surgeon's hands. To relax the need for surgical assistants and augment dexterity in surgery, it has been recently proposed to use a robot like a third arm that can be controlled seamlessly, independently from the natural arms, and work together with them. This report will develop and investigate this concept by implementing foot control of a robotic surgical arm. A novel passive haptic foot-machine interface system and analysis of its performances were introduced in this report. This interface using a parallel-serial hybrid structure with springs and force sensors, which allows intuitive control of a slave robotic arm with four degrees of freedom (dof). The elastic isometric design enables a user to control the interface system accurately and adaptively, with an enlarged sensing range breaking the physical restriction of the pedal size. A subject specific (independent component analysis, ICA) model is identified to map the surgeon's foot movements into kinematic parameters of the slave robotic arm. To validate the system and assess the performance it allows, 10 subjects carried out experiments to manipulate the foot-machine interface system in various movements. With these experimental data, the mapping models were built and verified. A comparison between different mapping models was made and analyzed proving the ICA algorithm is obviously dominant over other methods.
Liked
jechoi@andrew.cmu.edu
Development of an Intuitive Foot-Machine Interface for Robotic Surgery : The human-machine interface is of critical importance for master-slave control of the robotic system for surgery, in which current systems offer the control or two robotic arms teleoperated by the surgeon's hands. To relax the need for surgical assistants and augment dexterity in surgery, it has been recently proposed to use a robot like a third arm that can be controlled seamlessly, independently from the natural arms, and work together with them. This report will develop and investigate this concept by implementing foot control of a robotic surgical arm. A novel passive haptic foot-machine interface system and analysis of its performances were introduced in this report. This interface using a parallel-serial hybrid structure with springs and force sensors, which allows intuitive control of a slave robotic arm with four degrees of freedom (dof). The elastic isometric design enables a user to control the interface system accurately and adaptively, with an enlarged sensing range breaking the physical restriction of the pedal size. A subject specific (independent component analysis, ICA) model is identified to map the surgeon's foot movements into kinematic parameters of the slave robotic arm. To validate the system and assess the performance it allows, 10 subjects carried out experiments to manipulate the foot-machine interface system in various movements. With these experimental data, the mapping models were built and verified. A comparison between different mapping models was made and analyzed proving the ICA algorithm is obviously dominant over other methods.
1
jechoi@andrew.cmu.edu [SEP] Development of an Intuitive Foot-Machine Interface for Robotic Surgery : The human-machine interface is of critical importance for master-slave control of the robotic system for surgery, in which current systems offer the control or two robotic arms teleoperated by the surgeon's hands. To relax the need for surgical assistants and augment dexterity in surgery, it has been recently proposed to use a robot like a third arm that can be controlled seamlessly, independently from the natural arms, and work together with them. This report will develop and investigate this concept by implementing foot control of a robotic surgical arm. A novel passive haptic foot-machine interface system and analysis of its performances were introduced in this report. This interface using a parallel-serial hybrid structure with springs and force sensors, which allows intuitive control of a slave robotic arm with four degrees of freedom (dof). The elastic isometric design enables a user to control the interface system accurately and adaptively, with an enlarged sensing range breaking the physical restriction of the pedal size. A subject specific (independent component analysis, ICA) model is identified to map the surgeon's foot movements into kinematic parameters of the slave robotic arm. To validate the system and assess the performance it allows, 10 subjects carried out experiments to manipulate the foot-machine interface system in various movements. With these experimental data, the mapping models were built and verified. A comparison between different mapping models was made and analyzed proving the ICA algorithm is obviously dominant over other methods.
483
Domain Knowledge in Artificial Intelligence: Using Conceptual Modeling to Increase Machine Learning Accuracy and Explainability
Machine learning enables the extraction of useful information from large, diverse datasets. However, despite many successful applications, machine learning continues to suffer from performance and transparency issues. These challenges can be partially attributed to the limited use of domain knowledge by machine learning models. This research proposes using the domain knowledge represented in conceptual models to improve the preparation of the data used to train machine learning models. We develop and demonstrate a method, called the Conceptual Modeling for Machine Learning (CMML), which is comprised of guidelines for data preparation in machine learning and based on conceptual modeling constructs and principles. To assess the impact of CMML on machine learning outcomes, we first applied it to two real-world problems to evaluate its impact on model performance. We then solicited an assessment by data scientists on the applicability of the method. These results demonstrate the value of CMML for improving machine learning outcomes.
Liked
zrz@andrew.cmu.edu
Domain Knowledge in Artificial Intelligence: Using Conceptual Modeling to Increase Machine Learning Accuracy and Explainability : Machine learning enables the extraction of useful information from large, diverse datasets. However, despite many successful applications, machine learning continues to suffer from performance and transparency issues. These challenges can be partially attributed to the limited use of domain knowledge by machine learning models. This research proposes using the domain knowledge represented in conceptual models to improve the preparation of the data used to train machine learning models. We develop and demonstrate a method, called the Conceptual Modeling for Machine Learning (CMML), which is comprised of guidelines for data preparation in machine learning and based on conceptual modeling constructs and principles. To assess the impact of CMML on machine learning outcomes, we first applied it to two real-world problems to evaluate its impact on model performance. We then solicited an assessment by data scientists on the applicability of the method. These results demonstrate the value of CMML for improving machine learning outcomes.
1
zrz@andrew.cmu.edu [SEP] Domain Knowledge in Artificial Intelligence: Using Conceptual Modeling to Increase Machine Learning Accuracy and Explainability : Machine learning enables the extraction of useful information from large, diverse datasets. However, despite many successful applications, machine learning continues to suffer from performance and transparency issues. These challenges can be partially attributed to the limited use of domain knowledge by machine learning models. This research proposes using the domain knowledge represented in conceptual models to improve the preparation of the data used to train machine learning models. We develop and demonstrate a method, called the Conceptual Modeling for Machine Learning (CMML), which is comprised of guidelines for data preparation in machine learning and based on conceptual modeling constructs and principles. To assess the impact of CMML on machine learning outcomes, we first applied it to two real-world problems to evaluate its impact on model performance. We then solicited an assessment by data scientists on the applicability of the method. These results demonstrate the value of CMML for improving machine learning outcomes.
116
Neurofeedback-Driven 6-DOF Robotic Arm: Integration of Brain-Computer Interface with Arduino for Advanced Control
Brain computer interface (BCI) applications in robotics are becoming more famous and famous. People with disabilities are facing a real-time problem of doing simple activities such as grasping, handshaking etc. in order to aid with this problem, the use of brain signals to control actuators is showing a great importance. The Emotive Insight, a Brain-Computer Interface (BCI) device, is utilized in this project to collect brain signals and transform them into commands for controlling a robotic arm using an Arduino controller. The Emotive Insight captures brain signals, which are subsequently analyzed using Emotive software and connected with Arduino code. The HITI Brain software integrates these devices, allowing for smooth communication between brain activity and the robotic arm. This system demonstrates how brain impulses may be utilized to control external devices directly. The results showed that the system is applicable efficiently to robotic arms and also for prosthetic arms with Multi Degree of Freedom. In addition to that, the system can be used for other actuators such as bikes, mobile robots, wheelchairs etc.
Liked
jechoi@andrew.cmu.edu
Neurofeedback-Driven 6-DOF Robotic Arm: Integration of Brain-Computer Interface with Arduino for Advanced Control : Brain computer interface (BCI) applications in robotics are becoming more famous and famous. People with disabilities are facing a real-time problem of doing simple activities such as grasping, handshaking etc. in order to aid with this problem, the use of brain signals to control actuators is showing a great importance. The Emotive Insight, a Brain-Computer Interface (BCI) device, is utilized in this project to collect brain signals and transform them into commands for controlling a robotic arm using an Arduino controller. The Emotive Insight captures brain signals, which are subsequently analyzed using Emotive software and connected with Arduino code. The HITI Brain software integrates these devices, allowing for smooth communication between brain activity and the robotic arm. This system demonstrates how brain impulses may be utilized to control external devices directly. The results showed that the system is applicable efficiently to robotic arms and also for prosthetic arms with Multi Degree of Freedom. In addition to that, the system can be used for other actuators such as bikes, mobile robots, wheelchairs etc.
1
jechoi@andrew.cmu.edu [SEP] Neurofeedback-Driven 6-DOF Robotic Arm: Integration of Brain-Computer Interface with Arduino for Advanced Control : Brain computer interface (BCI) applications in robotics are becoming more famous and famous. People with disabilities are facing a real-time problem of doing simple activities such as grasping, handshaking etc. in order to aid with this problem, the use of brain signals to control actuators is showing a great importance. The Emotive Insight, a Brain-Computer Interface (BCI) device, is utilized in this project to collect brain signals and transform them into commands for controlling a robotic arm using an Arduino controller. The Emotive Insight captures brain signals, which are subsequently analyzed using Emotive software and connected with Arduino code. The HITI Brain software integrates these devices, allowing for smooth communication between brain activity and the robotic arm. This system demonstrates how brain impulses may be utilized to control external devices directly. The results showed that the system is applicable efficiently to robotic arms and also for prosthetic arms with Multi Degree of Freedom. In addition to that, the system can be used for other actuators such as bikes, mobile robots, wheelchairs etc.
448
One-Shot Dual-Arm Imitation Learning
We introduce One-Shot Dual-Arm Imitation Learning (ODIL), which enables dual-arm robots to learn precise and coordinated everyday tasks from just a single demonstration of the task. ODIL uses a new three-stage visual servoing (3-VS) method for precise alignment between the end-effector and target object, after which replay of the demonstration trajectory is sufficient to perform the task. This is achieved without requiring prior task or object knowledge, or additional data collection and training following the single demonstration. Furthermore, we propose a new dual-arm coordination paradigm for learning dual-arm tasks from a single demonstration. ODIL was tested on a real-world dual-arm robot, demonstrating state-of-the-art performance across six precise and coordinated tasks in both 4-DoF and 6-DoF settings, and showing robustness in the presence of distractor objects and partial occlusions. Videos are available at: https://www.robot-learning.uk/one-shot-dual-arm.
Liked
jechoi@andrew.cmu.edu
One-Shot Dual-Arm Imitation Learning : We introduce One-Shot Dual-Arm Imitation Learning (ODIL), which enables dual-arm robots to learn precise and coordinated everyday tasks from just a single demonstration of the task. ODIL uses a new three-stage visual servoing (3-VS) method for precise alignment between the end-effector and target object, after which replay of the demonstration trajectory is sufficient to perform the task. This is achieved without requiring prior task or object knowledge, or additional data collection and training following the single demonstration. Furthermore, we propose a new dual-arm coordination paradigm for learning dual-arm tasks from a single demonstration. ODIL was tested on a real-world dual-arm robot, demonstrating state-of-the-art performance across six precise and coordinated tasks in both 4-DoF and 6-DoF settings, and showing robustness in the presence of distractor objects and partial occlusions. Videos are available at: https://www.robot-learning.uk/one-shot-dual-arm.
1
jechoi@andrew.cmu.edu [SEP] One-Shot Dual-Arm Imitation Learning : We introduce One-Shot Dual-Arm Imitation Learning (ODIL), which enables dual-arm robots to learn precise and coordinated everyday tasks from just a single demonstration of the task. ODIL uses a new three-stage visual servoing (3-VS) method for precise alignment between the end-effector and target object, after which replay of the demonstration trajectory is sufficient to perform the task. This is achieved without requiring prior task or object knowledge, or additional data collection and training following the single demonstration. Furthermore, we propose a new dual-arm coordination paradigm for learning dual-arm tasks from a single demonstration. ODIL was tested on a real-world dual-arm robot, demonstrating state-of-the-art performance across six precise and coordinated tasks in both 4-DoF and 6-DoF settings, and showing robustness in the presence of distractor objects and partial occlusions. Videos are available at: https://www.robot-learning.uk/one-shot-dual-arm.
443
Harnessing The Multi-Stability Of Kresling Origami For Reconfigurable Articulation In Soft Robotic Arms
This study examines a biology-inspired approach of using reconfigurable articulation to reduce the control requirement for soft robotic arms. We construct a robotic arm by assembling Kresling origami modules that exhibit predictable bistability. Via switching between their two stable states, these origami modules can behave either like a flexible joint with low bending stiffness or like a stiff link with high stiffness, without requiring any continuous power supply. In this way, the robotic arm can exhibit pseudo-linkage kinematics with lower control requirements and improved motion accuracy. A unique advantage of using origami as the robotic arm skeleton is that its bending stiffness ratio between stable states is directly related to the underlying Kresling design. Therefore, we conduct extensive parametric analyses and experimental validations to identify the optimized Kresling pattern for articulation. The results indicate that a higher angle ratio, a smaller resting length at contracted stable state, and a large number of polygon sides can offer more significant and robust bending stiffness tuning. Based on this insight, we construct a proof-of-concept, tendon-driven robotic arm consisting of three modules, and show that it can exhibit the desired reconfigurable articulation behavior. Moreover, the deformations of this manipulator are consistent with kinematic model predictions, which validate the possibility of using simple controllers for such compliant robotic systems.
Liked
jechoi@andrew.cmu.edu
Harnessing The Multi-Stability Of Kresling Origami For Reconfigurable Articulation In Soft Robotic Arms : This study examines a biology-inspired approach of using reconfigurable articulation to reduce the control requirement for soft robotic arms. We construct a robotic arm by assembling Kresling origami modules that exhibit predictable bistability. Via switching between their two stable states, these origami modules can behave either like a flexible joint with low bending stiffness or like a stiff link with high stiffness, without requiring any continuous power supply. In this way, the robotic arm can exhibit pseudo-linkage kinematics with lower control requirements and improved motion accuracy. A unique advantage of using origami as the robotic arm skeleton is that its bending stiffness ratio between stable states is directly related to the underlying Kresling design. Therefore, we conduct extensive parametric analyses and experimental validations to identify the optimized Kresling pattern for articulation. The results indicate that a higher angle ratio, a smaller resting length at contracted stable state, and a large number of polygon sides can offer more significant and robust bending stiffness tuning. Based on this insight, we construct a proof-of-concept, tendon-driven robotic arm consisting of three modules, and show that it can exhibit the desired reconfigurable articulation behavior. Moreover, the deformations of this manipulator are consistent with kinematic model predictions, which validate the possibility of using simple controllers for such compliant robotic systems.
1
jechoi@andrew.cmu.edu [SEP] Harnessing The Multi-Stability Of Kresling Origami For Reconfigurable Articulation In Soft Robotic Arms : This study examines a biology-inspired approach of using reconfigurable articulation to reduce the control requirement for soft robotic arms. We construct a robotic arm by assembling Kresling origami modules that exhibit predictable bistability. Via switching between their two stable states, these origami modules can behave either like a flexible joint with low bending stiffness or like a stiff link with high stiffness, without requiring any continuous power supply. In this way, the robotic arm can exhibit pseudo-linkage kinematics with lower control requirements and improved motion accuracy. A unique advantage of using origami as the robotic arm skeleton is that its bending stiffness ratio between stable states is directly related to the underlying Kresling design. Therefore, we conduct extensive parametric analyses and experimental validations to identify the optimized Kresling pattern for articulation. The results indicate that a higher angle ratio, a smaller resting length at contracted stable state, and a large number of polygon sides can offer more significant and robust bending stiffness tuning. Based on this insight, we construct a proof-of-concept, tendon-driven robotic arm consisting of three modules, and show that it can exhibit the desired reconfigurable articulation behavior. Moreover, the deformations of this manipulator are consistent with kinematic model predictions, which validate the possibility of using simple controllers for such compliant robotic systems.
484
Survey on LiDAR Perception in Adverse Weather Conditions
Autonomous vehicles rely on a variety of sensors to gather information about their surrounding. The vehicle's behavior is planned based on the environment perception, making its reliability crucial for safety reasons. The active LiDAR sensor is able to create an accurate 3D representation of a scene, making it a valuable addition for environment perception for autonomous vehicles. Due to light scattering and occlusion, the LiDAR's performance change under adverse weather conditions like fog, snow or rain. This limitation recently fostered a large body of research on approaches to alleviate the decrease in perception performance. In this survey, we gathered, analyzed, and discussed different aspects on dealing with adverse weather conditions in LiDAR-based environment perception. We address topics such as the availability of appropriate data, raw point cloud processing and denoising, robust perception algorithms and sensor fusion to mitigate adverse weather induced shortcomings. We furthermore identify the most pressing gaps in the current literature and pinpoint promising research directions.
Disliked
zrz@andrew.cmu.edu
Survey on LiDAR Perception in Adverse Weather Conditions : Autonomous vehicles rely on a variety of sensors to gather information about their surrounding. The vehicle's behavior is planned based on the environment perception, making its reliability crucial for safety reasons. The active LiDAR sensor is able to create an accurate 3D representation of a scene, making it a valuable addition for environment perception for autonomous vehicles. Due to light scattering and occlusion, the LiDAR's performance change under adverse weather conditions like fog, snow or rain. This limitation recently fostered a large body of research on approaches to alleviate the decrease in perception performance. In this survey, we gathered, analyzed, and discussed different aspects on dealing with adverse weather conditions in LiDAR-based environment perception. We address topics such as the availability of appropriate data, raw point cloud processing and denoising, robust perception algorithms and sensor fusion to mitigate adverse weather induced shortcomings. We furthermore identify the most pressing gaps in the current literature and pinpoint promising research directions.
0
zrz@andrew.cmu.edu [SEP] Survey on LiDAR Perception in Adverse Weather Conditions : Autonomous vehicles rely on a variety of sensors to gather information about their surrounding. The vehicle's behavior is planned based on the environment perception, making its reliability crucial for safety reasons. The active LiDAR sensor is able to create an accurate 3D representation of a scene, making it a valuable addition for environment perception for autonomous vehicles. Due to light scattering and occlusion, the LiDAR's performance change under adverse weather conditions like fog, snow or rain. This limitation recently fostered a large body of research on approaches to alleviate the decrease in perception performance. In this survey, we gathered, analyzed, and discussed different aspects on dealing with adverse weather conditions in LiDAR-based environment perception. We address topics such as the availability of appropriate data, raw point cloud processing and denoising, robust perception algorithms and sensor fusion to mitigate adverse weather induced shortcomings. We furthermore identify the most pressing gaps in the current literature and pinpoint promising research directions.
334
Interruption-Aware Cooperative Perception for V2X Communication-Aided Autonomous Driving
Cooperative perception can significantly improve the perception performance of autonomous vehicles beyond the limited perception ability of individual vehicles by exchanging information with neighbor agents through V2X communication. However, most existing work assume ideal communication among agents, ignoring the significant and common \textit{interruption issues} caused by imperfect V2X communication, where cooperation agents can not receive cooperative messages successfully and thus fail to achieve cooperative perception, leading to safety risks. To fully reap the benefits of cooperative perception in practice, we propose V2X communication INterruption-aware COoperative Perception (V2X-INCOP), a cooperative perception system robust to communication interruption for V2X communication-aided autonomous driving, which leverages historical cooperation information to recover missing information due to the interruptions and alleviate the impact of the interruption issue. To achieve comprehensive recovery, we design a communication-adaptive multi-scale spatial-temporal prediction model to extract multi-scale spatial-temporal features based on V2X communication conditions and capture the most significant information for the prediction of the missing information. To further improve recovery performance, we adopt a knowledge distillation framework to give explicit and direct supervision to the prediction model and a curriculum learning strategy to stabilize the training of the model. Experiments on three public cooperative perception datasets demonstrate that the proposed method is effective in alleviating the impacts of communication interruption on cooperative perception.
Liked
zrz@andrew.cmu.edu
Interruption-Aware Cooperative Perception for V2X Communication-Aided Autonomous Driving : Cooperative perception can significantly improve the perception performance of autonomous vehicles beyond the limited perception ability of individual vehicles by exchanging information with neighbor agents through V2X communication. However, most existing work assume ideal communication among agents, ignoring the significant and common \textit{interruption issues} caused by imperfect V2X communication, where cooperation agents can not receive cooperative messages successfully and thus fail to achieve cooperative perception, leading to safety risks. To fully reap the benefits of cooperative perception in practice, we propose V2X communication INterruption-aware COoperative Perception (V2X-INCOP), a cooperative perception system robust to communication interruption for V2X communication-aided autonomous driving, which leverages historical cooperation information to recover missing information due to the interruptions and alleviate the impact of the interruption issue. To achieve comprehensive recovery, we design a communication-adaptive multi-scale spatial-temporal prediction model to extract multi-scale spatial-temporal features based on V2X communication conditions and capture the most significant information for the prediction of the missing information. To further improve recovery performance, we adopt a knowledge distillation framework to give explicit and direct supervision to the prediction model and a curriculum learning strategy to stabilize the training of the model. Experiments on three public cooperative perception datasets demonstrate that the proposed method is effective in alleviating the impacts of communication interruption on cooperative perception.
1
zrz@andrew.cmu.edu [SEP] Interruption-Aware Cooperative Perception for V2X Communication-Aided Autonomous Driving : Cooperative perception can significantly improve the perception performance of autonomous vehicles beyond the limited perception ability of individual vehicles by exchanging information with neighbor agents through V2X communication. However, most existing work assume ideal communication among agents, ignoring the significant and common \textit{interruption issues} caused by imperfect V2X communication, where cooperation agents can not receive cooperative messages successfully and thus fail to achieve cooperative perception, leading to safety risks. To fully reap the benefits of cooperative perception in practice, we propose V2X communication INterruption-aware COoperative Perception (V2X-INCOP), a cooperative perception system robust to communication interruption for V2X communication-aided autonomous driving, which leverages historical cooperation information to recover missing information due to the interruptions and alleviate the impact of the interruption issue. To achieve comprehensive recovery, we design a communication-adaptive multi-scale spatial-temporal prediction model to extract multi-scale spatial-temporal features based on V2X communication conditions and capture the most significant information for the prediction of the missing information. To further improve recovery performance, we adopt a knowledge distillation framework to give explicit and direct supervision to the prediction model and a curriculum learning strategy to stabilize the training of the model. Experiments on three public cooperative perception datasets demonstrate that the proposed method is effective in alleviating the impacts of communication interruption on cooperative perception.
285
Bridging Hard and Soft: Mechanical Metamaterials Enable Rigid Torque Transmission in Soft Robots
Torque and continuous rotation are fundamental methods of actuation and manipulation in rigid robots. Soft robot arms use soft materials and structures to mimic the passive compliance of biological arms that bend and extend. This use of compliance prevents soft arms from continuously transmitting and exerting torques to interact with their environment. Here, we show how relying on patterning structures instead of inherent material properties allows soft robotic arms to remain compliant while continuously transmitting torque to their environment. We demonstrate a soft robotic arm made from a pair of mechanical metamaterials that act as compliant constant-velocity joints. The joints are up to 52 times stiffer in torsion than bending and can bend up to 45{\deg}. This robot arm can continuously transmit torque while deforming in all other directions. The arm's mechanical design achieves high motion repeatability (0.4 mm and 0.1{\deg}) when tracking trajectories. We then trained a neural network to learn the inverse kinematics, enabling us to program the arm to complete tasks that are challenging for existing soft robots such as installing light bulbs, fastening bolts, and turning valves. The arm's passive compliance makes it safe around humans and provides a source of mechanical intelligence, enabling it to adapt to misalignment when manipulating objects. This work will bridge the gap between hard and soft robotics with applications in human assistance, warehouse automation, and extreme environments.
Liked
jechoi@andrew.cmu.edu
Bridging Hard and Soft: Mechanical Metamaterials Enable Rigid Torque Transmission in Soft Robots : Torque and continuous rotation are fundamental methods of actuation and manipulation in rigid robots. Soft robot arms use soft materials and structures to mimic the passive compliance of biological arms that bend and extend. This use of compliance prevents soft arms from continuously transmitting and exerting torques to interact with their environment. Here, we show how relying on patterning structures instead of inherent material properties allows soft robotic arms to remain compliant while continuously transmitting torque to their environment. We demonstrate a soft robotic arm made from a pair of mechanical metamaterials that act as compliant constant-velocity joints. The joints are up to 52 times stiffer in torsion than bending and can bend up to 45{\deg}. This robot arm can continuously transmit torque while deforming in all other directions. The arm's mechanical design achieves high motion repeatability (0.4 mm and 0.1{\deg}) when tracking trajectories. We then trained a neural network to learn the inverse kinematics, enabling us to program the arm to complete tasks that are challenging for existing soft robots such as installing light bulbs, fastening bolts, and turning valves. The arm's passive compliance makes it safe around humans and provides a source of mechanical intelligence, enabling it to adapt to misalignment when manipulating objects. This work will bridge the gap between hard and soft robotics with applications in human assistance, warehouse automation, and extreme environments.
1
jechoi@andrew.cmu.edu [SEP] Bridging Hard and Soft: Mechanical Metamaterials Enable Rigid Torque Transmission in Soft Robots : Torque and continuous rotation are fundamental methods of actuation and manipulation in rigid robots. Soft robot arms use soft materials and structures to mimic the passive compliance of biological arms that bend and extend. This use of compliance prevents soft arms from continuously transmitting and exerting torques to interact with their environment. Here, we show how relying on patterning structures instead of inherent material properties allows soft robotic arms to remain compliant while continuously transmitting torque to their environment. We demonstrate a soft robotic arm made from a pair of mechanical metamaterials that act as compliant constant-velocity joints. The joints are up to 52 times stiffer in torsion than bending and can bend up to 45{\deg}. This robot arm can continuously transmit torque while deforming in all other directions. The arm's mechanical design achieves high motion repeatability (0.4 mm and 0.1{\deg}) when tracking trajectories. We then trained a neural network to learn the inverse kinematics, enabling us to program the arm to complete tasks that are challenging for existing soft robots such as installing light bulbs, fastening bolts, and turning valves. The arm's passive compliance makes it safe around humans and provides a source of mechanical intelligence, enabling it to adapt to misalignment when manipulating objects. This work will bridge the gap between hard and soft robotics with applications in human assistance, warehouse automation, and extreme environments.
407
Autonomous Soil Collection in Environments With Heterogeneous Terrain
To autonomously collect soil in uncultivated terrain, robotic arms must distinguish between different amorphous materials and submerge themselves into the correct material. We develop a prototype that collects soil in heterogeneous terrain. If mounted to a mobile robot, it can be used to perform soil collection and analysis without human intervention. Unique among soil sampling robots, we use a general-purpose robotic arm rather than a soil core sampler.
Liked
jechoi@andrew.cmu.edu
Autonomous Soil Collection in Environments With Heterogeneous Terrain : To autonomously collect soil in uncultivated terrain, robotic arms must distinguish between different amorphous materials and submerge themselves into the correct material. We develop a prototype that collects soil in heterogeneous terrain. If mounted to a mobile robot, it can be used to perform soil collection and analysis without human intervention. Unique among soil sampling robots, we use a general-purpose robotic arm rather than a soil core sampler.
1
jechoi@andrew.cmu.edu [SEP] Autonomous Soil Collection in Environments With Heterogeneous Terrain : To autonomously collect soil in uncultivated terrain, robotic arms must distinguish between different amorphous materials and submerge themselves into the correct material. We develop a prototype that collects soil in heterogeneous terrain. If mounted to a mobile robot, it can be used to perform soil collection and analysis without human intervention. Unique among soil sampling robots, we use a general-purpose robotic arm rather than a soil core sampler.
519
Open Arms: Open-Source Arms, Hands & Control
Open Arms is a novel open-source platform of realistic human-like robotic hands and arms hardware with 28 Degree-of-Freedom (DoF), designed to extend the capabilities and accessibility of humanoid robotic grasping and manipulation. The Open Arms framework includes an open SDK and development environment, simulation tools, and application development tools to build and operate Open Arms. This paper describes these hands controls, sensing, mechanisms, aesthetic design, and manufacturing and their real-world applications with a teleoperated nursing robot. From 2015 to 2022, the authors have designed and established the manufacturing of Open Arms as a low-cost, high functionality robotic arms hardware and software framework to serve both humanoid robot applications and the urgent demand for low-cost prosthetics, as part of the Hanson Robotics Sophia Robot platform. Using the techniques of consumer product manufacturing, we set out to define modular, low-cost techniques for approximating the dexterity and sensitivity of human hands. To demonstrate the dexterity and control of our hands, we present a Generative Grasping Residual CNN (GGR-CNN) model that can generate robust antipodal grasps from input images of various objects in real-time speeds (22ms). We achieved state-of-the-art accuracy of 92.4% using our model architecture on a standard Cornell Grasping Dataset, which contains a diverse set of household objects.
Liked
jechoi@andrew.cmu.edu
Open Arms: Open-Source Arms, Hands & Control : Open Arms is a novel open-source platform of realistic human-like robotic hands and arms hardware with 28 Degree-of-Freedom (DoF), designed to extend the capabilities and accessibility of humanoid robotic grasping and manipulation. The Open Arms framework includes an open SDK and development environment, simulation tools, and application development tools to build and operate Open Arms. This paper describes these hands controls, sensing, mechanisms, aesthetic design, and manufacturing and their real-world applications with a teleoperated nursing robot. From 2015 to 2022, the authors have designed and established the manufacturing of Open Arms as a low-cost, high functionality robotic arms hardware and software framework to serve both humanoid robot applications and the urgent demand for low-cost prosthetics, as part of the Hanson Robotics Sophia Robot platform. Using the techniques of consumer product manufacturing, we set out to define modular, low-cost techniques for approximating the dexterity and sensitivity of human hands. To demonstrate the dexterity and control of our hands, we present a Generative Grasping Residual CNN (GGR-CNN) model that can generate robust antipodal grasps from input images of various objects in real-time speeds (22ms). We achieved state-of-the-art accuracy of 92.4% using our model architecture on a standard Cornell Grasping Dataset, which contains a diverse set of household objects.
1
jechoi@andrew.cmu.edu [SEP] Open Arms: Open-Source Arms, Hands & Control : Open Arms is a novel open-source platform of realistic human-like robotic hands and arms hardware with 28 Degree-of-Freedom (DoF), designed to extend the capabilities and accessibility of humanoid robotic grasping and manipulation. The Open Arms framework includes an open SDK and development environment, simulation tools, and application development tools to build and operate Open Arms. This paper describes these hands controls, sensing, mechanisms, aesthetic design, and manufacturing and their real-world applications with a teleoperated nursing robot. From 2015 to 2022, the authors have designed and established the manufacturing of Open Arms as a low-cost, high functionality robotic arms hardware and software framework to serve both humanoid robot applications and the urgent demand for low-cost prosthetics, as part of the Hanson Robotics Sophia Robot platform. Using the techniques of consumer product manufacturing, we set out to define modular, low-cost techniques for approximating the dexterity and sensitivity of human hands. To demonstrate the dexterity and control of our hands, we present a Generative Grasping Residual CNN (GGR-CNN) model that can generate robust antipodal grasps from input images of various objects in real-time speeds (22ms). We achieved state-of-the-art accuracy of 92.4% using our model architecture on a standard Cornell Grasping Dataset, which contains a diverse set of household objects.
434
Deep frequency principle towards understanding why deeper learning is faster
Understanding the effect of depth in deep learning is a critical problem. In this work, we utilize the Fourier analysis to empirically provide a promising mechanism to understand why feedforward deeper learning is faster. To this end, we separate a deep neural network, trained by normal stochastic gradient descent, into two parts during analysis, i.e., a pre-condition component and a learning component, in which the output of the pre-condition one is the input of the learning one. We use a filtering method to characterize the frequency distribution of a high-dimensional function. Based on experiments of deep networks and real dataset, we propose a deep frequency principle, that is, the effective target function for a deeper hidden layer biases towards lower frequency during the training. Therefore, the learning component effectively learns a lower frequency function if the pre-condition component has more layers. Due to the well-studied frequency principle, i.e., deep neural networks learn lower frequency functions faster, the deep frequency principle provides a reasonable explanation to why deeper learning is faster. We believe these empirical studies would be valuable for future theoretical studies of the effect of depth in deep learning.
Liked
zrz@andrew.cmu.edu
Deep frequency principle towards understanding why deeper learning is faster : Understanding the effect of depth in deep learning is a critical problem. In this work, we utilize the Fourier analysis to empirically provide a promising mechanism to understand why feedforward deeper learning is faster. To this end, we separate a deep neural network, trained by normal stochastic gradient descent, into two parts during analysis, i.e., a pre-condition component and a learning component, in which the output of the pre-condition one is the input of the learning one. We use a filtering method to characterize the frequency distribution of a high-dimensional function. Based on experiments of deep networks and real dataset, we propose a deep frequency principle, that is, the effective target function for a deeper hidden layer biases towards lower frequency during the training. Therefore, the learning component effectively learns a lower frequency function if the pre-condition component has more layers. Due to the well-studied frequency principle, i.e., deep neural networks learn lower frequency functions faster, the deep frequency principle provides a reasonable explanation to why deeper learning is faster. We believe these empirical studies would be valuable for future theoretical studies of the effect of depth in deep learning.
1
zrz@andrew.cmu.edu [SEP] Deep frequency principle towards understanding why deeper learning is faster : Understanding the effect of depth in deep learning is a critical problem. In this work, we utilize the Fourier analysis to empirically provide a promising mechanism to understand why feedforward deeper learning is faster. To this end, we separate a deep neural network, trained by normal stochastic gradient descent, into two parts during analysis, i.e., a pre-condition component and a learning component, in which the output of the pre-condition one is the input of the learning one. We use a filtering method to characterize the frequency distribution of a high-dimensional function. Based on experiments of deep networks and real dataset, we propose a deep frequency principle, that is, the effective target function for a deeper hidden layer biases towards lower frequency during the training. Therefore, the learning component effectively learns a lower frequency function if the pre-condition component has more layers. Due to the well-studied frequency principle, i.e., deep neural networks learn lower frequency functions faster, the deep frequency principle provides a reasonable explanation to why deeper learning is faster. We believe these empirical studies would be valuable for future theoretical studies of the effect of depth in deep learning.
194
RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins
In the rapidly advancing field of robotics, dual-arm coordination and complex object manipulation are essential capabilities for developing advanced autonomous systems. However, the scarcity of diverse, high-quality demonstration data and real-world-aligned evaluation benchmarks severely limits such development. To address this, we introduce RoboTwin, a generative digital twin framework that uses 3D generative foundation models and large language models to produce diverse expert datasets and provide a real-world-aligned evaluation platform for dual-arm robotic tasks. Specifically, RoboTwin creates varied digital twins of objects from single 2D images, generating realistic and interactive scenarios. It also introduces a spatial relation-aware code generation framework that combines object annotations with large language models to break down tasks, determine spatial constraints, and generate precise robotic movement code. Our framework offers a comprehensive benchmark with both simulated and real-world data, enabling standardized evaluation and better alignment between simulated training and real-world performance. We validated our approach using the open-source COBOT Magic Robot platform. Policies pre-trained on RoboTwin-generated data and fine-tuned with limited real-world samples demonstrate significant potential for enhancing dual-arm robotic manipulation systems by improving success rates by over 70% for single-arm tasks and over 40% for dual-arm tasks compared to models trained solely on real-world data.
Liked
jechoi@andrew.cmu.edu
RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins : In the rapidly advancing field of robotics, dual-arm coordination and complex object manipulation are essential capabilities for developing advanced autonomous systems. However, the scarcity of diverse, high-quality demonstration data and real-world-aligned evaluation benchmarks severely limits such development. To address this, we introduce RoboTwin, a generative digital twin framework that uses 3D generative foundation models and large language models to produce diverse expert datasets and provide a real-world-aligned evaluation platform for dual-arm robotic tasks. Specifically, RoboTwin creates varied digital twins of objects from single 2D images, generating realistic and interactive scenarios. It also introduces a spatial relation-aware code generation framework that combines object annotations with large language models to break down tasks, determine spatial constraints, and generate precise robotic movement code. Our framework offers a comprehensive benchmark with both simulated and real-world data, enabling standardized evaluation and better alignment between simulated training and real-world performance. We validated our approach using the open-source COBOT Magic Robot platform. Policies pre-trained on RoboTwin-generated data and fine-tuned with limited real-world samples demonstrate significant potential for enhancing dual-arm robotic manipulation systems by improving success rates by over 70% for single-arm tasks and over 40% for dual-arm tasks compared to models trained solely on real-world data.
1
jechoi@andrew.cmu.edu [SEP] RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins : In the rapidly advancing field of robotics, dual-arm coordination and complex object manipulation are essential capabilities for developing advanced autonomous systems. However, the scarcity of diverse, high-quality demonstration data and real-world-aligned evaluation benchmarks severely limits such development. To address this, we introduce RoboTwin, a generative digital twin framework that uses 3D generative foundation models and large language models to produce diverse expert datasets and provide a real-world-aligned evaluation platform for dual-arm robotic tasks. Specifically, RoboTwin creates varied digital twins of objects from single 2D images, generating realistic and interactive scenarios. It also introduces a spatial relation-aware code generation framework that combines object annotations with large language models to break down tasks, determine spatial constraints, and generate precise robotic movement code. Our framework offers a comprehensive benchmark with both simulated and real-world data, enabling standardized evaluation and better alignment between simulated training and real-world performance. We validated our approach using the open-source COBOT Magic Robot platform. Policies pre-trained on RoboTwin-generated data and fine-tuned with limited real-world samples demonstrate significant potential for enhancing dual-arm robotic manipulation systems by improving success rates by over 70% for single-arm tasks and over 40% for dual-arm tasks compared to models trained solely on real-world data.
511
Planning to Build Soma Blocks Using a Dual-arm Robot
This paper presents a planner that can automatically find an optimal assembly sequence for a dual-arm robot to assemble the soma blocks. The planner uses the mesh model of objects and the final state of the assembly to generate all possible assembly sequence and evaluate the optimal assembly sequence by considering the stability, graspability, assemblability, as well as the need for a second arm. Especially, the need for a second arm is considered when supports from worktables and other workpieces are not enough to produce a stable assembly. The planner will refer to an assisting grasp to additionally hold and support the unstable components so that the robot can further assemble new workpieces and finally reach a stable state. The output of the planner is the optimal assembly orders, candidate grasps, assembly directions, and the assisting grasps if any. The output of the planner can be used to guide a dual-arm robot to perform the assembly task. The planner is verified in both simulations and real-world executions.
Liked
jechoi@andrew.cmu.edu
Planning to Build Soma Blocks Using a Dual-arm Robot : This paper presents a planner that can automatically find an optimal assembly sequence for a dual-arm robot to assemble the soma blocks. The planner uses the mesh model of objects and the final state of the assembly to generate all possible assembly sequence and evaluate the optimal assembly sequence by considering the stability, graspability, assemblability, as well as the need for a second arm. Especially, the need for a second arm is considered when supports from worktables and other workpieces are not enough to produce a stable assembly. The planner will refer to an assisting grasp to additionally hold and support the unstable components so that the robot can further assemble new workpieces and finally reach a stable state. The output of the planner is the optimal assembly orders, candidate grasps, assembly directions, and the assisting grasps if any. The output of the planner can be used to guide a dual-arm robot to perform the assembly task. The planner is verified in both simulations and real-world executions.
1
jechoi@andrew.cmu.edu [SEP] Planning to Build Soma Blocks Using a Dual-arm Robot : This paper presents a planner that can automatically find an optimal assembly sequence for a dual-arm robot to assemble the soma blocks. The planner uses the mesh model of objects and the final state of the assembly to generate all possible assembly sequence and evaluate the optimal assembly sequence by considering the stability, graspability, assemblability, as well as the need for a second arm. Especially, the need for a second arm is considered when supports from worktables and other workpieces are not enough to produce a stable assembly. The planner will refer to an assisting grasp to additionally hold and support the unstable components so that the robot can further assemble new workpieces and finally reach a stable state. The output of the planner is the optimal assembly orders, candidate grasps, assembly directions, and the assisting grasps if any. The output of the planner can be used to guide a dual-arm robot to perform the assembly task. The planner is verified in both simulations and real-world executions.
497
Cybathlon -- Legged Mobile Assistance for Quadriplegics
Assistance robots are the future for people who need daily care due to limited mobility or being wheelchair-bound. Current solutions of attaching robotic arms to motorized wheelchairs only provide limited additional mobility at the cost of increased size. We present a mouth joystick control interface, augmented with voice commands, for an independent quadrupedal assistance robot with an arm. We validate and showcase our system in the Cybathlon Challenges February 2024 Assistance Robot Race, where we solve four everyday tasks in record time, winning first place. Our system remains generic and sets the basis for a platform that could help and provide independence in the everyday lives of people in wheelchairs.
Disliked
jechoi@andrew.cmu.edu
Cybathlon -- Legged Mobile Assistance for Quadriplegics : Assistance robots are the future for people who need daily care due to limited mobility or being wheelchair-bound. Current solutions of attaching robotic arms to motorized wheelchairs only provide limited additional mobility at the cost of increased size. We present a mouth joystick control interface, augmented with voice commands, for an independent quadrupedal assistance robot with an arm. We validate and showcase our system in the Cybathlon Challenges February 2024 Assistance Robot Race, where we solve four everyday tasks in record time, winning first place. Our system remains generic and sets the basis for a platform that could help and provide independence in the everyday lives of people in wheelchairs.
0
jechoi@andrew.cmu.edu [SEP] Cybathlon -- Legged Mobile Assistance for Quadriplegics : Assistance robots are the future for people who need daily care due to limited mobility or being wheelchair-bound. Current solutions of attaching robotic arms to motorized wheelchairs only provide limited additional mobility at the cost of increased size. We present a mouth joystick control interface, augmented with voice commands, for an independent quadrupedal assistance robot with an arm. We validate and showcase our system in the Cybathlon Challenges February 2024 Assistance Robot Race, where we solve four everyday tasks in record time, winning first place. Our system remains generic and sets the basis for a platform that could help and provide independence in the everyday lives of people in wheelchairs.
524
RGB-D Robotic Pose Estimation For a Servicing Robotic Arm
A large number of robotic and human-assisted missions to the Moon and Mars are forecast. NASA's efforts to learn about the geology and makeup of these celestial bodies rely heavily on the use of robotic arms. The safety and redundancy aspects will be crucial when humans will be working alongside the robotic explorers. Additionally, robotic arms are crucial to satellite servicing and planned orbit debris mitigation missions. The goal of this work is to create a custom Computer Vision (CV) based Artificial Neural Network (ANN) that would be able to rapidly identify the posture of a 7 Degree of Freedom (DoF) robotic arm from a single (RGB-D) image - just like humans can easily identify if an arm is pointing in some general direction. The Sawyer robotic arm is used for developing and training this intelligent algorithm. Since Sawyer's joint space spans 7 dimensions, it is an insurmountable task to cover the entire joint configuration space. In this work, orthogonal arrays are used, similar to the Taguchi method, to efficiently span the joint space with the minimal number of training images. This ``optimally'' generated database is used to train the custom ANN and its degree of accuracy is on average equal to twice the smallest joint displacement step used for database generation. A pre-trained ANN will be useful for estimating the postures of robotic manipulators used on space stations, spacecraft, and rovers as an auxiliary tool or for contingency plans.
Liked
jechoi@andrew.cmu.edu
RGB-D Robotic Pose Estimation For a Servicing Robotic Arm : A large number of robotic and human-assisted missions to the Moon and Mars are forecast. NASA's efforts to learn about the geology and makeup of these celestial bodies rely heavily on the use of robotic arms. The safety and redundancy aspects will be crucial when humans will be working alongside the robotic explorers. Additionally, robotic arms are crucial to satellite servicing and planned orbit debris mitigation missions. The goal of this work is to create a custom Computer Vision (CV) based Artificial Neural Network (ANN) that would be able to rapidly identify the posture of a 7 Degree of Freedom (DoF) robotic arm from a single (RGB-D) image - just like humans can easily identify if an arm is pointing in some general direction. The Sawyer robotic arm is used for developing and training this intelligent algorithm. Since Sawyer's joint space spans 7 dimensions, it is an insurmountable task to cover the entire joint configuration space. In this work, orthogonal arrays are used, similar to the Taguchi method, to efficiently span the joint space with the minimal number of training images. This ``optimally'' generated database is used to train the custom ANN and its degree of accuracy is on average equal to twice the smallest joint displacement step used for database generation. A pre-trained ANN will be useful for estimating the postures of robotic manipulators used on space stations, spacecraft, and rovers as an auxiliary tool or for contingency plans.
1
jechoi@andrew.cmu.edu [SEP] RGB-D Robotic Pose Estimation For a Servicing Robotic Arm : A large number of robotic and human-assisted missions to the Moon and Mars are forecast. NASA's efforts to learn about the geology and makeup of these celestial bodies rely heavily on the use of robotic arms. The safety and redundancy aspects will be crucial when humans will be working alongside the robotic explorers. Additionally, robotic arms are crucial to satellite servicing and planned orbit debris mitigation missions. The goal of this work is to create a custom Computer Vision (CV) based Artificial Neural Network (ANN) that would be able to rapidly identify the posture of a 7 Degree of Freedom (DoF) robotic arm from a single (RGB-D) image - just like humans can easily identify if an arm is pointing in some general direction. The Sawyer robotic arm is used for developing and training this intelligent algorithm. Since Sawyer's joint space spans 7 dimensions, it is an insurmountable task to cover the entire joint configuration space. In this work, orthogonal arrays are used, similar to the Taguchi method, to efficiently span the joint space with the minimal number of training images. This ``optimally'' generated database is used to train the custom ANN and its degree of accuracy is on average equal to twice the smallest joint displacement step used for database generation. A pre-trained ANN will be useful for estimating the postures of robotic manipulators used on space stations, spacecraft, and rovers as an auxiliary tool or for contingency plans.
449
Optimal Multi-Manipulator Arm Placement for Maximal Dexterity during Robotics Surgery
Robot arm placements are oftentimes a limitation in surgical preoperative procedures, relying on trained staff to evaluate and decide on the optimal positions for the arms. Given new and different patient anatomies, it can be challenging to make an informed choice, leading to more frequently colliding arms or limited manipulator workspaces. In this paper, we develop a method to generate the optimal manipulator base positions for the multi-port da Vinci surgical system that minimizes self-collision and environment-collision, and maximizes the surgeon's reachability inside the patient. Scoring functions are defined for each criterion so that they may be optimized over. Since for multi-manipulator setups, a large number of free parameters are available to adjust the base positioning of each arm, a challenge becomes how one can expediently assess possible setups. We thus also propose methods that perform fast queries of each measure with the use of a proxy collision-checker. We then develop an optimization method to determine the optimal position using the scoring functions. We evaluate the optimality of the base positions for the robot arms on canonical trajectories, and show that the solution yielded by the optimization program can satisfy each criterion. The metrics and optimization strategy are generalizable to other surgical robotic platforms so that patient-side manipulator positioning may be optimized and solved.
Liked
jechoi@andrew.cmu.edu
Optimal Multi-Manipulator Arm Placement for Maximal Dexterity during Robotics Surgery : Robot arm placements are oftentimes a limitation in surgical preoperative procedures, relying on trained staff to evaluate and decide on the optimal positions for the arms. Given new and different patient anatomies, it can be challenging to make an informed choice, leading to more frequently colliding arms or limited manipulator workspaces. In this paper, we develop a method to generate the optimal manipulator base positions for the multi-port da Vinci surgical system that minimizes self-collision and environment-collision, and maximizes the surgeon's reachability inside the patient. Scoring functions are defined for each criterion so that they may be optimized over. Since for multi-manipulator setups, a large number of free parameters are available to adjust the base positioning of each arm, a challenge becomes how one can expediently assess possible setups. We thus also propose methods that perform fast queries of each measure with the use of a proxy collision-checker. We then develop an optimization method to determine the optimal position using the scoring functions. We evaluate the optimality of the base positions for the robot arms on canonical trajectories, and show that the solution yielded by the optimization program can satisfy each criterion. The metrics and optimization strategy are generalizable to other surgical robotic platforms so that patient-side manipulator positioning may be optimized and solved.
1
jechoi@andrew.cmu.edu [SEP] Optimal Multi-Manipulator Arm Placement for Maximal Dexterity during Robotics Surgery : Robot arm placements are oftentimes a limitation in surgical preoperative procedures, relying on trained staff to evaluate and decide on the optimal positions for the arms. Given new and different patient anatomies, it can be challenging to make an informed choice, leading to more frequently colliding arms or limited manipulator workspaces. In this paper, we develop a method to generate the optimal manipulator base positions for the multi-port da Vinci surgical system that minimizes self-collision and environment-collision, and maximizes the surgeon's reachability inside the patient. Scoring functions are defined for each criterion so that they may be optimized over. Since for multi-manipulator setups, a large number of free parameters are available to adjust the base positioning of each arm, a challenge becomes how one can expediently assess possible setups. We thus also propose methods that perform fast queries of each measure with the use of a proxy collision-checker. We then develop an optimization method to determine the optimal position using the scoring functions. We evaluate the optimality of the base positions for the robot arms on canonical trajectories, and show that the solution yielded by the optimization program can satisfy each criterion. The metrics and optimization strategy are generalizable to other surgical robotic platforms so that patient-side manipulator positioning may be optimized and solved.
553
Developing and Comparing Single-arm and Dual-arm Regrasp
The goal of this paper is to develop efficient regrasp algorithms for single-arm and dual-arm regrasp and compares the performance of single-arm and dual-arm regrasp by running the two algorithms thousands of times. We focus on pick-and-place regrasp which reorients an object from one placement to another by using a sequence of pick-ups and place-downs. After analyzing the simulation results, we find dual-arm regrasp is not necessarily better than single-arm regrasp: Dual-arm regrasp is flexible. When the two hands can grasp the object with good clearance, dual-arm regrasp is better and has higher successful rate than single-arm regrasp. However, dual-arm regrasp suffers from geometric constraints caused by the two arms. When the grasps overlap, dual-arm regrasp is bad. Developers need to sample grasps with high density to reduce overlapping. This leads to exploded combinatorics in previous methods, but is possible with the algorithms presented in this paper. Following the results, practitioners may choose single-arm or dual-arm robots by considering the object shapes and grasps. Meanwhile, they can reduce overlapping and implement practical dual-arm regrasp by using the presented algorithms.
Liked
jechoi@andrew.cmu.edu
Developing and Comparing Single-arm and Dual-arm Regrasp : The goal of this paper is to develop efficient regrasp algorithms for single-arm and dual-arm regrasp and compares the performance of single-arm and dual-arm regrasp by running the two algorithms thousands of times. We focus on pick-and-place regrasp which reorients an object from one placement to another by using a sequence of pick-ups and place-downs. After analyzing the simulation results, we find dual-arm regrasp is not necessarily better than single-arm regrasp: Dual-arm regrasp is flexible. When the two hands can grasp the object with good clearance, dual-arm regrasp is better and has higher successful rate than single-arm regrasp. However, dual-arm regrasp suffers from geometric constraints caused by the two arms. When the grasps overlap, dual-arm regrasp is bad. Developers need to sample grasps with high density to reduce overlapping. This leads to exploded combinatorics in previous methods, but is possible with the algorithms presented in this paper. Following the results, practitioners may choose single-arm or dual-arm robots by considering the object shapes and grasps. Meanwhile, they can reduce overlapping and implement practical dual-arm regrasp by using the presented algorithms.
1
jechoi@andrew.cmu.edu [SEP] Developing and Comparing Single-arm and Dual-arm Regrasp : The goal of this paper is to develop efficient regrasp algorithms for single-arm and dual-arm regrasp and compares the performance of single-arm and dual-arm regrasp by running the two algorithms thousands of times. We focus on pick-and-place regrasp which reorients an object from one placement to another by using a sequence of pick-ups and place-downs. After analyzing the simulation results, we find dual-arm regrasp is not necessarily better than single-arm regrasp: Dual-arm regrasp is flexible. When the two hands can grasp the object with good clearance, dual-arm regrasp is better and has higher successful rate than single-arm regrasp. However, dual-arm regrasp suffers from geometric constraints caused by the two arms. When the grasps overlap, dual-arm regrasp is bad. Developers need to sample grasps with high density to reduce overlapping. This leads to exploded combinatorics in previous methods, but is possible with the algorithms presented in this paper. Following the results, practitioners may choose single-arm or dual-arm robots by considering the object shapes and grasps. Meanwhile, they can reduce overlapping and implement practical dual-arm regrasp by using the presented algorithms.
8
Words or Vision: Do Vision-Language Models Have Blind Faith in Text?
Vision-Language Models (VLMs) excel in integrating visual and textual information for vision-centric tasks, but their handling of inconsistencies between modalities is underexplored. We investigate VLMs' modality preferences when faced with visual data and varied textual inputs in vision-centered settings. By introducing textual variations to four vision-centric tasks and evaluating ten Vision-Language Models (VLMs), we discover a \emph{``blind faith in text''} phenomenon: VLMs disproportionately trust textual data over visual data when inconsistencies arise, leading to significant performance drops under corrupted text and raising safety concerns. We analyze factors influencing this text bias, including instruction prompts, language model size, text relevance, token order, and the interplay between visual and textual certainty. While certain factors, such as scaling up the language model size, slightly mitigate text bias, others like token order can exacerbate it due to positional biases inherited from language models. To address this issue, we explore supervised fine-tuning with text augmentation and demonstrate its effectiveness in reducing text bias. Additionally, we provide a theoretical analysis suggesting that the blind faith in text phenomenon may stem from an imbalance of pure text and multi-modal data during training. Our findings highlight the need for balanced training and careful consideration of modality interactions in VLMs to enhance their robustness and reliability in handling multi-modal data inconsistencies.
Disliked
zrz@andrew.cmu.edu
Words or Vision: Do Vision-Language Models Have Blind Faith in Text? : Vision-Language Models (VLMs) excel in integrating visual and textual information for vision-centric tasks, but their handling of inconsistencies between modalities is underexplored. We investigate VLMs' modality preferences when faced with visual data and varied textual inputs in vision-centered settings. By introducing textual variations to four vision-centric tasks and evaluating ten Vision-Language Models (VLMs), we discover a \emph{``blind faith in text''} phenomenon: VLMs disproportionately trust textual data over visual data when inconsistencies arise, leading to significant performance drops under corrupted text and raising safety concerns. We analyze factors influencing this text bias, including instruction prompts, language model size, text relevance, token order, and the interplay between visual and textual certainty. While certain factors, such as scaling up the language model size, slightly mitigate text bias, others like token order can exacerbate it due to positional biases inherited from language models. To address this issue, we explore supervised fine-tuning with text augmentation and demonstrate its effectiveness in reducing text bias. Additionally, we provide a theoretical analysis suggesting that the blind faith in text phenomenon may stem from an imbalance of pure text and multi-modal data during training. Our findings highlight the need for balanced training and careful consideration of modality interactions in VLMs to enhance their robustness and reliability in handling multi-modal data inconsistencies.
0
zrz@andrew.cmu.edu [SEP] Words or Vision: Do Vision-Language Models Have Blind Faith in Text? : Vision-Language Models (VLMs) excel in integrating visual and textual information for vision-centric tasks, but their handling of inconsistencies between modalities is underexplored. We investigate VLMs' modality preferences when faced with visual data and varied textual inputs in vision-centered settings. By introducing textual variations to four vision-centric tasks and evaluating ten Vision-Language Models (VLMs), we discover a \emph{``blind faith in text''} phenomenon: VLMs disproportionately trust textual data over visual data when inconsistencies arise, leading to significant performance drops under corrupted text and raising safety concerns. We analyze factors influencing this text bias, including instruction prompts, language model size, text relevance, token order, and the interplay between visual and textual certainty. While certain factors, such as scaling up the language model size, slightly mitigate text bias, others like token order can exacerbate it due to positional biases inherited from language models. To address this issue, we explore supervised fine-tuning with text augmentation and demonstrate its effectiveness in reducing text bias. Additionally, we provide a theoretical analysis suggesting that the blind faith in text phenomenon may stem from an imbalance of pure text and multi-modal data during training. Our findings highlight the need for balanced training and careful consideration of modality interactions in VLMs to enhance their robustness and reliability in handling multi-modal data inconsistencies.
351
3D Hand-Eye Calibration for Collaborative Robot Arm: Look at Robot Base Once
Hand-eye calibration is a common problem in the field of collaborative robotics, involving the determination of the transformation matrix between the visual sensor and the robot flange to enable vision-based robotic tasks. However, this process typically requires multiple movements of the robot arm and an external calibration object, making it both time-consuming and inconvenient, especially in scenarios where frequent recalibration is necessary. In this work, we extend our previous method which eliminates the need for external calibration objects such as a chessboard. We propose a generic dataset generation approach for point cloud registration, focusing on aligning the robot base point cloud with the scanned data. Furthermore, a more detailed simulation study is conducted involving several different collaborative robot arms, followed by real-world experiments in an industrial setting. Our improved method is simulated and evaluated using a total of 14 robotic arms from 9 different brands, including KUKA, Universal Robots, UFACTORY, and Franka Emika, all of which are widely used in the field of collaborative robotics. Physical experiments demonstrate that our extended approach achieves performance comparable to existing commercial hand-eye calibration solutions, while completing the entire calibration procedure in just a few seconds. In addition, we provide a user-friendly hand-eye calibration solution, with the code publicly available at github.com/leihui6/LRBO.
Liked
jechoi@andrew.cmu.edu
3D Hand-Eye Calibration for Collaborative Robot Arm: Look at Robot Base Once : Hand-eye calibration is a common problem in the field of collaborative robotics, involving the determination of the transformation matrix between the visual sensor and the robot flange to enable vision-based robotic tasks. However, this process typically requires multiple movements of the robot arm and an external calibration object, making it both time-consuming and inconvenient, especially in scenarios where frequent recalibration is necessary. In this work, we extend our previous method which eliminates the need for external calibration objects such as a chessboard. We propose a generic dataset generation approach for point cloud registration, focusing on aligning the robot base point cloud with the scanned data. Furthermore, a more detailed simulation study is conducted involving several different collaborative robot arms, followed by real-world experiments in an industrial setting. Our improved method is simulated and evaluated using a total of 14 robotic arms from 9 different brands, including KUKA, Universal Robots, UFACTORY, and Franka Emika, all of which are widely used in the field of collaborative robotics. Physical experiments demonstrate that our extended approach achieves performance comparable to existing commercial hand-eye calibration solutions, while completing the entire calibration procedure in just a few seconds. In addition, we provide a user-friendly hand-eye calibration solution, with the code publicly available at github.com/leihui6/LRBO.
1
jechoi@andrew.cmu.edu [SEP] 3D Hand-Eye Calibration for Collaborative Robot Arm: Look at Robot Base Once : Hand-eye calibration is a common problem in the field of collaborative robotics, involving the determination of the transformation matrix between the visual sensor and the robot flange to enable vision-based robotic tasks. However, this process typically requires multiple movements of the robot arm and an external calibration object, making it both time-consuming and inconvenient, especially in scenarios where frequent recalibration is necessary. In this work, we extend our previous method which eliminates the need for external calibration objects such as a chessboard. We propose a generic dataset generation approach for point cloud registration, focusing on aligning the robot base point cloud with the scanned data. Furthermore, a more detailed simulation study is conducted involving several different collaborative robot arms, followed by real-world experiments in an industrial setting. Our improved method is simulated and evaluated using a total of 14 robotic arms from 9 different brands, including KUKA, Universal Robots, UFACTORY, and Franka Emika, all of which are widely used in the field of collaborative robotics. Physical experiments demonstrate that our extended approach achieves performance comparable to existing commercial hand-eye calibration solutions, while completing the entire calibration procedure in just a few seconds. In addition, we provide a user-friendly hand-eye calibration solution, with the code publicly available at github.com/leihui6/LRBO.
495
Beneficial and Harmful Explanatory Machine Learning
Given the recent successes of Deep Learning in AI there has been increased interest in the role and need for explanations in machine learned theories. A distinct notion in this context is that of Michie's definition of Ultra-Strong Machine Learning (USML). USML is demonstrated by a measurable increase in human performance of a task following provision to the human of a symbolic machine learned theory for task performance. A recent paper demonstrates the beneficial effect of a machine learned logic theory for a classification task, yet no existing work to our knowledge has examined the potential harmfulness of machine's involvement for human comprehension during learning. This paper investigates the explanatory effects of a machine learned theory in the context of simple two person games and proposes a framework for identifying the harmfulness of machine explanations based on the Cognitive Science literature. The approach involves a cognitive window consisting of two quantifiable bounds and it is supported by empirical evidence collected from human trials. Our quantitative and qualitative results indicate that human learning aided by a symbolic machine learned theory which satisfies a cognitive window has achieved significantly higher performance than human self learning. Results also demonstrate that human learning aided by a symbolic machine learned theory that fails to satisfy this window leads to significantly worse performance than unaided human learning.
Liked
zrz@andrew.cmu.edu
Beneficial and Harmful Explanatory Machine Learning : Given the recent successes of Deep Learning in AI there has been increased interest in the role and need for explanations in machine learned theories. A distinct notion in this context is that of Michie's definition of Ultra-Strong Machine Learning (USML). USML is demonstrated by a measurable increase in human performance of a task following provision to the human of a symbolic machine learned theory for task performance. A recent paper demonstrates the beneficial effect of a machine learned logic theory for a classification task, yet no existing work to our knowledge has examined the potential harmfulness of machine's involvement for human comprehension during learning. This paper investigates the explanatory effects of a machine learned theory in the context of simple two person games and proposes a framework for identifying the harmfulness of machine explanations based on the Cognitive Science literature. The approach involves a cognitive window consisting of two quantifiable bounds and it is supported by empirical evidence collected from human trials. Our quantitative and qualitative results indicate that human learning aided by a symbolic machine learned theory which satisfies a cognitive window has achieved significantly higher performance than human self learning. Results also demonstrate that human learning aided by a symbolic machine learned theory that fails to satisfy this window leads to significantly worse performance than unaided human learning.
1
zrz@andrew.cmu.edu [SEP] Beneficial and Harmful Explanatory Machine Learning : Given the recent successes of Deep Learning in AI there has been increased interest in the role and need for explanations in machine learned theories. A distinct notion in this context is that of Michie's definition of Ultra-Strong Machine Learning (USML). USML is demonstrated by a measurable increase in human performance of a task following provision to the human of a symbolic machine learned theory for task performance. A recent paper demonstrates the beneficial effect of a machine learned logic theory for a classification task, yet no existing work to our knowledge has examined the potential harmfulness of machine's involvement for human comprehension during learning. This paper investigates the explanatory effects of a machine learned theory in the context of simple two person games and proposes a framework for identifying the harmfulness of machine explanations based on the Cognitive Science literature. The approach involves a cognitive window consisting of two quantifiable bounds and it is supported by empirical evidence collected from human trials. Our quantitative and qualitative results indicate that human learning aided by a symbolic machine learned theory which satisfies a cognitive window has achieved significantly higher performance than human self learning. Results also demonstrate that human learning aided by a symbolic machine learned theory that fails to satisfy this window leads to significantly worse performance than unaided human learning.
126
Optimal Trajectory Planning for Flexible Robots with Large Deformation
Robot arms with lighter weight can reduce unnecessary energy consumption which is desirable in robotic industry. However, lightweight arms undergo undesirable elastic deformation. In this paper, the planar motion of a lightweight flexible arm is investigated. In order to obtain a precise mathematical model, the axial displacement and nonlinear curvature of flexible arm arising from large bending deformation is taken into consideration. An in-extensional condition, the axial displacement is related to transverse displacement of the flexible beam, is applied. This leads to a robotic model with three rigid modes and one elastic mode. The elastic mode depends on time and position. An Assume Mode Method is used to remove the spatial dependence. The governing equations is derived using Lagrange Method. The effects of nonlinear terms due to the large deformation, gravity, and tip-mass are considered. Control inputs include forces and moment exerted at the joint between slider and arm (see Fig. 1). The conventional computed torque control laws cannot stabilize the system, since there are not as many control inputs as states of the system. A Particle Swarm Optimization (PSO) technique is then used to obtain a suitable trajectory with the aim of minimizing excitations of the elastic mode. Two methods are considered for generating a trajectory function, either to use a three-layer Artificial Neural Network (ANN) or to use spline interpolation. A sliding mode control strategy is proposed in which the sliding surfaces include elastic mode in order to guarantee robustness. The simulations show that the three-layer ANN technique provides arbitrary small settling time, and also the optimization algorithm converges faster and generates smooth trajectories unlike spline function technique.
Liked
jechoi@andrew.cmu.edu
Optimal Trajectory Planning for Flexible Robots with Large Deformation : Robot arms with lighter weight can reduce unnecessary energy consumption which is desirable in robotic industry. However, lightweight arms undergo undesirable elastic deformation. In this paper, the planar motion of a lightweight flexible arm is investigated. In order to obtain a precise mathematical model, the axial displacement and nonlinear curvature of flexible arm arising from large bending deformation is taken into consideration. An in-extensional condition, the axial displacement is related to transverse displacement of the flexible beam, is applied. This leads to a robotic model with three rigid modes and one elastic mode. The elastic mode depends on time and position. An Assume Mode Method is used to remove the spatial dependence. The governing equations is derived using Lagrange Method. The effects of nonlinear terms due to the large deformation, gravity, and tip-mass are considered. Control inputs include forces and moment exerted at the joint between slider and arm (see Fig. 1). The conventional computed torque control laws cannot stabilize the system, since there are not as many control inputs as states of the system. A Particle Swarm Optimization (PSO) technique is then used to obtain a suitable trajectory with the aim of minimizing excitations of the elastic mode. Two methods are considered for generating a trajectory function, either to use a three-layer Artificial Neural Network (ANN) or to use spline interpolation. A sliding mode control strategy is proposed in which the sliding surfaces include elastic mode in order to guarantee robustness. The simulations show that the three-layer ANN technique provides arbitrary small settling time, and also the optimization algorithm converges faster and generates smooth trajectories unlike spline function technique.
1
jechoi@andrew.cmu.edu [SEP] Optimal Trajectory Planning for Flexible Robots with Large Deformation : Robot arms with lighter weight can reduce unnecessary energy consumption which is desirable in robotic industry. However, lightweight arms undergo undesirable elastic deformation. In this paper, the planar motion of a lightweight flexible arm is investigated. In order to obtain a precise mathematical model, the axial displacement and nonlinear curvature of flexible arm arising from large bending deformation is taken into consideration. An in-extensional condition, the axial displacement is related to transverse displacement of the flexible beam, is applied. This leads to a robotic model with three rigid modes and one elastic mode. The elastic mode depends on time and position. An Assume Mode Method is used to remove the spatial dependence. The governing equations is derived using Lagrange Method. The effects of nonlinear terms due to the large deformation, gravity, and tip-mass are considered. Control inputs include forces and moment exerted at the joint between slider and arm (see Fig. 1). The conventional computed torque control laws cannot stabilize the system, since there are not as many control inputs as states of the system. A Particle Swarm Optimization (PSO) technique is then used to obtain a suitable trajectory with the aim of minimizing excitations of the elastic mode. Two methods are considered for generating a trajectory function, either to use a three-layer Artificial Neural Network (ANN) or to use spline interpolation. A sliding mode control strategy is proposed in which the sliding surfaces include elastic mode in order to guarantee robustness. The simulations show that the three-layer ANN technique provides arbitrary small settling time, and also the optimization algorithm converges faster and generates smooth trajectories unlike spline function technique.
572
Application of deep reinforcement learning for Indian stock trading automation
In stock trading, feature extraction and trading strategy design are the two important tasks to achieve long-term benefits using machine learning techniques. Several methods have been proposed to design trading strategy by acquiring trading signals to maximize the rewards. In the present paper the theory of deep reinforcement learning is applied for stock trading strategy and investment decisions to Indian markets. The experiments are performed systematically with three classical Deep Reinforcement Learning models Deep Q-Network, Double Deep Q-Network and Dueling Double Deep Q-Network on ten Indian stock datasets. The performance of the models are evaluated and comparison is made.
Liked
zrz@andrew.cmu.edu
Application of deep reinforcement learning for Indian stock trading automation : In stock trading, feature extraction and trading strategy design are the two important tasks to achieve long-term benefits using machine learning techniques. Several methods have been proposed to design trading strategy by acquiring trading signals to maximize the rewards. In the present paper the theory of deep reinforcement learning is applied for stock trading strategy and investment decisions to Indian markets. The experiments are performed systematically with three classical Deep Reinforcement Learning models Deep Q-Network, Double Deep Q-Network and Dueling Double Deep Q-Network on ten Indian stock datasets. The performance of the models are evaluated and comparison is made.
1
zrz@andrew.cmu.edu [SEP] Application of deep reinforcement learning for Indian stock trading automation : In stock trading, feature extraction and trading strategy design are the two important tasks to achieve long-term benefits using machine learning techniques. Several methods have been proposed to design trading strategy by acquiring trading signals to maximize the rewards. In the present paper the theory of deep reinforcement learning is applied for stock trading strategy and investment decisions to Indian markets. The experiments are performed systematically with three classical Deep Reinforcement Learning models Deep Q-Network, Double Deep Q-Network and Dueling Double Deep Q-Network on ten Indian stock datasets. The performance of the models are evaluated and comparison is made.
219
Dynamic Movement Primitive based Motion Retargeting for Dual-Arm Sign Language Motions
We aim to develop an efficient programming method for equipping service robots with the skill of performing sign language motions. This paper addresses the problem of transferring complex dual-arm sign language motions characterized by the coordination among arms and hands from human to robot, which is seldom considered in previous studies of motion retargeting techniques. In this paper, we propose a novel motion retargeting method that leverages graph optimization and Dynamic Movement Primitives (DMPs) for this problem. We employ DMPs in a leader-follower manner to parameterize the original trajectories while preserving motion rhythm and relative movements between human body parts, and adopt a three-step optimization procedure to find deformed trajectories for robot motion planning while ensuring feasibility for robot execution. Experimental results of several Chinese Sign Language (CSL) motions have been successfully performed on ABB's YuMi dual-arm collaborative robot (14-DOF) with two 6-DOF Inspire-Robotics' multi-fingered hands, a system with 26 DOFs in total.
Liked
jechoi@andrew.cmu.edu
Dynamic Movement Primitive based Motion Retargeting for Dual-Arm Sign Language Motions : We aim to develop an efficient programming method for equipping service robots with the skill of performing sign language motions. This paper addresses the problem of transferring complex dual-arm sign language motions characterized by the coordination among arms and hands from human to robot, which is seldom considered in previous studies of motion retargeting techniques. In this paper, we propose a novel motion retargeting method that leverages graph optimization and Dynamic Movement Primitives (DMPs) for this problem. We employ DMPs in a leader-follower manner to parameterize the original trajectories while preserving motion rhythm and relative movements between human body parts, and adopt a three-step optimization procedure to find deformed trajectories for robot motion planning while ensuring feasibility for robot execution. Experimental results of several Chinese Sign Language (CSL) motions have been successfully performed on ABB's YuMi dual-arm collaborative robot (14-DOF) with two 6-DOF Inspire-Robotics' multi-fingered hands, a system with 26 DOFs in total.
1
jechoi@andrew.cmu.edu [SEP] Dynamic Movement Primitive based Motion Retargeting for Dual-Arm Sign Language Motions : We aim to develop an efficient programming method for equipping service robots with the skill of performing sign language motions. This paper addresses the problem of transferring complex dual-arm sign language motions characterized by the coordination among arms and hands from human to robot, which is seldom considered in previous studies of motion retargeting techniques. In this paper, we propose a novel motion retargeting method that leverages graph optimization and Dynamic Movement Primitives (DMPs) for this problem. We employ DMPs in a leader-follower manner to parameterize the original trajectories while preserving motion rhythm and relative movements between human body parts, and adopt a three-step optimization procedure to find deformed trajectories for robot motion planning while ensuring feasibility for robot execution. Experimental results of several Chinese Sign Language (CSL) motions have been successfully performed on ABB's YuMi dual-arm collaborative robot (14-DOF) with two 6-DOF Inspire-Robotics' multi-fingered hands, a system with 26 DOFs in total.
489
An Overview of Deep Semi-Supervised Learning
Deep neural networks demonstrated their ability to provide remarkable performances on a wide range of supervised learning tasks (e.g., image classification) when trained on extensive collections of labeled data (e.g., ImageNet). However, creating such large datasets requires a considerable amount of resources, time, and effort. Such resources may not be available in many practical cases, limiting the adoption and the application of many deep learning methods. In a search for more data-efficient deep learning methods to overcome the need for large annotated datasets, there is a rising research interest in semi-supervised learning and its applications to deep neural networks to reduce the amount of labeled data required, by either developing novel methods or adopting existing semi-supervised learning frameworks for a deep learning setting. In this paper, we provide a comprehensive overview of deep semi-supervised learning, starting with an introduction to the field, followed by a summarization of the dominant semi-supervised approaches in deep learning.
Liked
zrz@andrew.cmu.edu
An Overview of Deep Semi-Supervised Learning : Deep neural networks demonstrated their ability to provide remarkable performances on a wide range of supervised learning tasks (e.g., image classification) when trained on extensive collections of labeled data (e.g., ImageNet). However, creating such large datasets requires a considerable amount of resources, time, and effort. Such resources may not be available in many practical cases, limiting the adoption and the application of many deep learning methods. In a search for more data-efficient deep learning methods to overcome the need for large annotated datasets, there is a rising research interest in semi-supervised learning and its applications to deep neural networks to reduce the amount of labeled data required, by either developing novel methods or adopting existing semi-supervised learning frameworks for a deep learning setting. In this paper, we provide a comprehensive overview of deep semi-supervised learning, starting with an introduction to the field, followed by a summarization of the dominant semi-supervised approaches in deep learning.
1
zrz@andrew.cmu.edu [SEP] An Overview of Deep Semi-Supervised Learning : Deep neural networks demonstrated their ability to provide remarkable performances on a wide range of supervised learning tasks (e.g., image classification) when trained on extensive collections of labeled data (e.g., ImageNet). However, creating such large datasets requires a considerable amount of resources, time, and effort. Such resources may not be available in many practical cases, limiting the adoption and the application of many deep learning methods. In a search for more data-efficient deep learning methods to overcome the need for large annotated datasets, there is a rising research interest in semi-supervised learning and its applications to deep neural networks to reduce the amount of labeled data required, by either developing novel methods or adopting existing semi-supervised learning frameworks for a deep learning setting. In this paper, we provide a comprehensive overview of deep semi-supervised learning, starting with an introduction to the field, followed by a summarization of the dominant semi-supervised approaches in deep learning.
200
Task Oriented Video Coding: A Survey
Video coding technology has been continuously improved for higher compression ratio with higher resolution. However, the state-of-the-art video coding standards, such as H.265/HEVC and Versatile Video Coding, are still designed with the assumption the compressed video will be watched by humans. With the tremendous advance and maturation of deep neural networks in solving computer vision tasks, more and more videos are directly analyzed by deep neural networks without humans' involvement. Such a conventional design for video coding standard is not optimal when the compressed video is used by computer vision applications. While the human visual system is consistently sensitive to the content with high contrast, the impact of pixels on computer vision algorithms is driven by specific computer vision tasks. In this paper, we explore and summarize recent progress on computer vision task oriented video coding and emerging video coding standard, Video Coding for Machines.
Disliked
zrz@andrew.cmu.edu
Task Oriented Video Coding: A Survey : Video coding technology has been continuously improved for higher compression ratio with higher resolution. However, the state-of-the-art video coding standards, such as H.265/HEVC and Versatile Video Coding, are still designed with the assumption the compressed video will be watched by humans. With the tremendous advance and maturation of deep neural networks in solving computer vision tasks, more and more videos are directly analyzed by deep neural networks without humans' involvement. Such a conventional design for video coding standard is not optimal when the compressed video is used by computer vision applications. While the human visual system is consistently sensitive to the content with high contrast, the impact of pixels on computer vision algorithms is driven by specific computer vision tasks. In this paper, we explore and summarize recent progress on computer vision task oriented video coding and emerging video coding standard, Video Coding for Machines.
0
zrz@andrew.cmu.edu [SEP] Task Oriented Video Coding: A Survey : Video coding technology has been continuously improved for higher compression ratio with higher resolution. However, the state-of-the-art video coding standards, such as H.265/HEVC and Versatile Video Coding, are still designed with the assumption the compressed video will be watched by humans. With the tremendous advance and maturation of deep neural networks in solving computer vision tasks, more and more videos are directly analyzed by deep neural networks without humans' involvement. Such a conventional design for video coding standard is not optimal when the compressed video is used by computer vision applications. While the human visual system is consistently sensitive to the content with high contrast, the impact of pixels on computer vision algorithms is driven by specific computer vision tasks. In this paper, we explore and summarize recent progress on computer vision task oriented video coding and emerging video coding standard, Video Coding for Machines.
354
How Developers Iterate on Machine Learning Workflows -- A Survey of the Applied Machine Learning Literature
Machine learning workflow development is anecdotally regarded to be an iterative process of trial-and-error with humans-in-the-loop. However, we are not aware of quantitative evidence corroborating this popular belief. A quantitative characterization of iteration can serve as a benchmark for machine learning workflow development in practice, and can aid the development of human-in-the-loop machine learning systems. To this end, we conduct a small-scale survey of the applied machine learning literature from five distinct application domains. We collect and distill statistics on the role of iteration within machine learning workflow development, and report preliminary trends and insights from our investigation, as a starting point towards this benchmark. Based on our findings, we finally describe desiderata for effective and versatile human-in-the-loop machine learning systems that can cater to users in diverse domains.
Liked
zrz@andrew.cmu.edu
How Developers Iterate on Machine Learning Workflows -- A Survey of the Applied Machine Learning Literature : Machine learning workflow development is anecdotally regarded to be an iterative process of trial-and-error with humans-in-the-loop. However, we are not aware of quantitative evidence corroborating this popular belief. A quantitative characterization of iteration can serve as a benchmark for machine learning workflow development in practice, and can aid the development of human-in-the-loop machine learning systems. To this end, we conduct a small-scale survey of the applied machine learning literature from five distinct application domains. We collect and distill statistics on the role of iteration within machine learning workflow development, and report preliminary trends and insights from our investigation, as a starting point towards this benchmark. Based on our findings, we finally describe desiderata for effective and versatile human-in-the-loop machine learning systems that can cater to users in diverse domains.
1
zrz@andrew.cmu.edu [SEP] How Developers Iterate on Machine Learning Workflows -- A Survey of the Applied Machine Learning Literature : Machine learning workflow development is anecdotally regarded to be an iterative process of trial-and-error with humans-in-the-loop. However, we are not aware of quantitative evidence corroborating this popular belief. A quantitative characterization of iteration can serve as a benchmark for machine learning workflow development in practice, and can aid the development of human-in-the-loop machine learning systems. To this end, we conduct a small-scale survey of the applied machine learning literature from five distinct application domains. We collect and distill statistics on the role of iteration within machine learning workflow development, and report preliminary trends and insights from our investigation, as a starting point towards this benchmark. Based on our findings, we finally describe desiderata for effective and versatile human-in-the-loop machine learning systems that can cater to users in diverse domains.
108
Learning Task-aware Robust Deep Learning Systems
Many works demonstrate that deep learning system is vulnerable to adversarial attack. A deep learning system consists of two parts: the deep learning task and the deep model. Nowadays, most existing works investigate the impact of the deep model on robustness of deep learning systems, ignoring the impact of the learning task. In this paper, we adopt the binary and interval label encoding strategy to redefine the classification task and design corresponding loss to improve robustness of the deep learning system. Our method can be viewed as improving the robustness of deep learning systems from both the learning task and deep model. Experimental results demonstrate that our learning task-aware method is much more robust than traditional classification while retaining the accuracy.
Liked
zrz@andrew.cmu.edu
Learning Task-aware Robust Deep Learning Systems : Many works demonstrate that deep learning system is vulnerable to adversarial attack. A deep learning system consists of two parts: the deep learning task and the deep model. Nowadays, most existing works investigate the impact of the deep model on robustness of deep learning systems, ignoring the impact of the learning task. In this paper, we adopt the binary and interval label encoding strategy to redefine the classification task and design corresponding loss to improve robustness of the deep learning system. Our method can be viewed as improving the robustness of deep learning systems from both the learning task and deep model. Experimental results demonstrate that our learning task-aware method is much more robust than traditional classification while retaining the accuracy.
1
zrz@andrew.cmu.edu [SEP] Learning Task-aware Robust Deep Learning Systems : Many works demonstrate that deep learning system is vulnerable to adversarial attack. A deep learning system consists of two parts: the deep learning task and the deep model. Nowadays, most existing works investigate the impact of the deep model on robustness of deep learning systems, ignoring the impact of the learning task. In this paper, we adopt the binary and interval label encoding strategy to redefine the classification task and design corresponding loss to improve robustness of the deep learning system. Our method can be viewed as improving the robustness of deep learning systems from both the learning task and deep model. Experimental results demonstrate that our learning task-aware method is much more robust than traditional classification while retaining the accuracy.
162
To New Beginnings: A Survey of Unified Perception in Autonomous Vehicle Software
Autonomous vehicle perception typically relies on modular pipelines that decompose the task into detection, tracking, and prediction. While interpretable, these pipelines suffer from error accumulation and limited inter-task synergy. Unified perception has emerged as a promising paradigm that integrates these sub-tasks within a shared architecture, potentially improving robustness, contextual reasoning, and efficiency while retaining interpretable outputs. In this survey, we provide a comprehensive overview of unified perception, introducing a holistic and systemic taxonomy that categorizes methods along task integration, tracking formulation, and representation flow. We define three paradigms -Early, Late, and Full Unified Perception- and systematically review existing methods, their architectures, training strategies, datasets used, and open-source availability, while highlighting future research directions. This work establishes the first comprehensive framework for understanding and advancing unified perception, consolidates fragmented efforts, and guides future research toward more robust, generalizable, and interpretable perception.
Disliked
zrz@andrew.cmu.edu
To New Beginnings: A Survey of Unified Perception in Autonomous Vehicle Software : Autonomous vehicle perception typically relies on modular pipelines that decompose the task into detection, tracking, and prediction. While interpretable, these pipelines suffer from error accumulation and limited inter-task synergy. Unified perception has emerged as a promising paradigm that integrates these sub-tasks within a shared architecture, potentially improving robustness, contextual reasoning, and efficiency while retaining interpretable outputs. In this survey, we provide a comprehensive overview of unified perception, introducing a holistic and systemic taxonomy that categorizes methods along task integration, tracking formulation, and representation flow. We define three paradigms -Early, Late, and Full Unified Perception- and systematically review existing methods, their architectures, training strategies, datasets used, and open-source availability, while highlighting future research directions. This work establishes the first comprehensive framework for understanding and advancing unified perception, consolidates fragmented efforts, and guides future research toward more robust, generalizable, and interpretable perception.
0
zrz@andrew.cmu.edu [SEP] To New Beginnings: A Survey of Unified Perception in Autonomous Vehicle Software : Autonomous vehicle perception typically relies on modular pipelines that decompose the task into detection, tracking, and prediction. While interpretable, these pipelines suffer from error accumulation and limited inter-task synergy. Unified perception has emerged as a promising paradigm that integrates these sub-tasks within a shared architecture, potentially improving robustness, contextual reasoning, and efficiency while retaining interpretable outputs. In this survey, we provide a comprehensive overview of unified perception, introducing a holistic and systemic taxonomy that categorizes methods along task integration, tracking formulation, and representation flow. We define three paradigms -Early, Late, and Full Unified Perception- and systematically review existing methods, their architectures, training strategies, datasets used, and open-source availability, while highlighting future research directions. This work establishes the first comprehensive framework for understanding and advancing unified perception, consolidates fragmented efforts, and guides future research toward more robust, generalizable, and interpretable perception.
335
Advances in Hybrid Modular Climbing Robots: Design Principles and Refinement Strategies
This paper explores the design strategies for hybrid pole- or trunk-climbing robots, focusing on methods to inform design decisions and assess metrics such as adaptability and performance. A wheeled-grasping hybrid robot with modular, tendon-driven grasping arms and a wheeled drive system mounted on a turret was developed to climb columns of varying diameters. Here, the key innovation is the underactuated arms that can be adjusted to different column sizes by adding or removing modular linkages, though the robot also features capabilities like self-locking (the ability of the robot to stay on the column by friction without power), autonomous grasping, and rotation around the column axis. Mathematical models describe conditions for self-locking and vertical climbing. Experimental results demonstrate the robot's efficacy in climbing and self-locking, validating the proposed models and highlighting the potential for fully automated solutions in industrial applications. This work provides a comprehensive framework for evaluating and designing hybrid climbing robots, contributing to advancements in autonomous robotics for environments where climbing tall structures is critical.
Liked
jechoi@andrew.cmu.edu
Advances in Hybrid Modular Climbing Robots: Design Principles and Refinement Strategies : This paper explores the design strategies for hybrid pole- or trunk-climbing robots, focusing on methods to inform design decisions and assess metrics such as adaptability and performance. A wheeled-grasping hybrid robot with modular, tendon-driven grasping arms and a wheeled drive system mounted on a turret was developed to climb columns of varying diameters. Here, the key innovation is the underactuated arms that can be adjusted to different column sizes by adding or removing modular linkages, though the robot also features capabilities like self-locking (the ability of the robot to stay on the column by friction without power), autonomous grasping, and rotation around the column axis. Mathematical models describe conditions for self-locking and vertical climbing. Experimental results demonstrate the robot's efficacy in climbing and self-locking, validating the proposed models and highlighting the potential for fully automated solutions in industrial applications. This work provides a comprehensive framework for evaluating and designing hybrid climbing robots, contributing to advancements in autonomous robotics for environments where climbing tall structures is critical.
1
jechoi@andrew.cmu.edu [SEP] Advances in Hybrid Modular Climbing Robots: Design Principles and Refinement Strategies : This paper explores the design strategies for hybrid pole- or trunk-climbing robots, focusing on methods to inform design decisions and assess metrics such as adaptability and performance. A wheeled-grasping hybrid robot with modular, tendon-driven grasping arms and a wheeled drive system mounted on a turret was developed to climb columns of varying diameters. Here, the key innovation is the underactuated arms that can be adjusted to different column sizes by adding or removing modular linkages, though the robot also features capabilities like self-locking (the ability of the robot to stay on the column by friction without power), autonomous grasping, and rotation around the column axis. Mathematical models describe conditions for self-locking and vertical climbing. Experimental results demonstrate the robot's efficacy in climbing and self-locking, validating the proposed models and highlighting the potential for fully automated solutions in industrial applications. This work provides a comprehensive framework for evaluating and designing hybrid climbing robots, contributing to advancements in autonomous robotics for environments where climbing tall structures is critical.
564
Distributed Deep Reinforcement Learning: A Survey and A Multi-Player Multi-Agent Learning Toolbox
With the breakthrough of AlphaGo, deep reinforcement learning becomes a recognized technique for solving sequential decision-making problems. Despite its reputation, data inefficiency caused by its trial and error learning mechanism makes deep reinforcement learning hard to be practical in a wide range of areas. Plenty of methods have been developed for sample efficient deep reinforcement learning, such as environment modeling, experience transfer, and distributed modifications, amongst which, distributed deep reinforcement learning has shown its potential in various applications, such as human-computer gaming, and intelligent transportation. In this paper, we conclude the state of this exciting field, by comparing the classical distributed deep reinforcement learning methods, and studying important components to achieve efficient distributed learning, covering single player single agent distributed deep reinforcement learning to the most complex multiple players multiple agents distributed deep reinforcement learning. Furthermore, we review recently released toolboxes that help to realize distributed deep reinforcement learning without many modifications of their non-distributed versions. By analyzing their strengths and weaknesses, a multi-player multi-agent distributed deep reinforcement learning toolbox is developed and released, which is further validated on Wargame, a complex environment, showing usability of the proposed toolbox for multiple players and multiple agents distributed deep reinforcement learning under complex games. Finally, we try to point out challenges and future trends, hoping this brief review can provide a guide or a spark for researchers who are interested in distributed deep reinforcement learning.
Disliked
zrz@andrew.cmu.edu
Distributed Deep Reinforcement Learning: A Survey and A Multi-Player Multi-Agent Learning Toolbox : With the breakthrough of AlphaGo, deep reinforcement learning becomes a recognized technique for solving sequential decision-making problems. Despite its reputation, data inefficiency caused by its trial and error learning mechanism makes deep reinforcement learning hard to be practical in a wide range of areas. Plenty of methods have been developed for sample efficient deep reinforcement learning, such as environment modeling, experience transfer, and distributed modifications, amongst which, distributed deep reinforcement learning has shown its potential in various applications, such as human-computer gaming, and intelligent transportation. In this paper, we conclude the state of this exciting field, by comparing the classical distributed deep reinforcement learning methods, and studying important components to achieve efficient distributed learning, covering single player single agent distributed deep reinforcement learning to the most complex multiple players multiple agents distributed deep reinforcement learning. Furthermore, we review recently released toolboxes that help to realize distributed deep reinforcement learning without many modifications of their non-distributed versions. By analyzing their strengths and weaknesses, a multi-player multi-agent distributed deep reinforcement learning toolbox is developed and released, which is further validated on Wargame, a complex environment, showing usability of the proposed toolbox for multiple players and multiple agents distributed deep reinforcement learning under complex games. Finally, we try to point out challenges and future trends, hoping this brief review can provide a guide or a spark for researchers who are interested in distributed deep reinforcement learning.
0
zrz@andrew.cmu.edu [SEP] Distributed Deep Reinforcement Learning: A Survey and A Multi-Player Multi-Agent Learning Toolbox : With the breakthrough of AlphaGo, deep reinforcement learning becomes a recognized technique for solving sequential decision-making problems. Despite its reputation, data inefficiency caused by its trial and error learning mechanism makes deep reinforcement learning hard to be practical in a wide range of areas. Plenty of methods have been developed for sample efficient deep reinforcement learning, such as environment modeling, experience transfer, and distributed modifications, amongst which, distributed deep reinforcement learning has shown its potential in various applications, such as human-computer gaming, and intelligent transportation. In this paper, we conclude the state of this exciting field, by comparing the classical distributed deep reinforcement learning methods, and studying important components to achieve efficient distributed learning, covering single player single agent distributed deep reinforcement learning to the most complex multiple players multiple agents distributed deep reinforcement learning. Furthermore, we review recently released toolboxes that help to realize distributed deep reinforcement learning without many modifications of their non-distributed versions. By analyzing their strengths and weaknesses, a multi-player multi-agent distributed deep reinforcement learning toolbox is developed and released, which is further validated on Wargame, a complex environment, showing usability of the proposed toolbox for multiple players and multiple agents distributed deep reinforcement learning under complex games. Finally, we try to point out challenges and future trends, hoping this brief review can provide a guide or a spark for researchers who are interested in distributed deep reinforcement learning.
179
Evaluation Challenges for Geospatial ML
As geospatial machine learning models and maps derived from their predictions are increasingly used for downstream analyses in science and policy, it is imperative to evaluate their accuracy and applicability. Geospatial machine learning has key distinctions from other learning paradigms, and as such, the correct way to measure performance of spatial machine learning outputs has been a topic of debate. In this paper, I delineate unique challenges of model evaluation for geospatial machine learning with global or remotely sensed datasets, culminating in concrete takeaways to improve evaluations of geospatial model performance.
Disliked
zrz@andrew.cmu.edu
Evaluation Challenges for Geospatial ML : As geospatial machine learning models and maps derived from their predictions are increasingly used for downstream analyses in science and policy, it is imperative to evaluate their accuracy and applicability. Geospatial machine learning has key distinctions from other learning paradigms, and as such, the correct way to measure performance of spatial machine learning outputs has been a topic of debate. In this paper, I delineate unique challenges of model evaluation for geospatial machine learning with global or remotely sensed datasets, culminating in concrete takeaways to improve evaluations of geospatial model performance.
0
zrz@andrew.cmu.edu [SEP] Evaluation Challenges for Geospatial ML : As geospatial machine learning models and maps derived from their predictions are increasingly used for downstream analyses in science and policy, it is imperative to evaluate their accuracy and applicability. Geospatial machine learning has key distinctions from other learning paradigms, and as such, the correct way to measure performance of spatial machine learning outputs has been a topic of debate. In this paper, I delineate unique challenges of model evaluation for geospatial machine learning with global or remotely sensed datasets, culminating in concrete takeaways to improve evaluations of geospatial model performance.
52
A Mobile Quad-Arm Robot ARMS: Wheeled-Legged Tripedal Locomotion and Quad-Arm Loco-Manipulation
This article proposes a mobile quad-arm robot: ARMS, which unifies wheeled-legged tripedal locomotion, wheeled locomotion, and quad-arm loco-manipulation. ARMS's four arms have different mechanisms and are partially designed to be general-purpose arms for the hybrid locomotion and loco-manipulation. One three-degree-of-freedom (DOF) arm has an active wheel, which is used for wheeled-legged tripedal walking and wheeled driving with passive wheels attached to the torso. Two three-DOF general-purpose arms are series elastic and used for wheeled-legged tripedal walking, object grasping, and manipulation. The upper two-DOF arm is used for manipulation only; its position and orientation are determined by coordinating all arms. Each motor is controlled by an angle controller and trajectory modification with angle, angular velocity, angular acceleration, and torque constraints. ARMS was verified with seven experiments involving joint control, wheeled-legged locomotion, wheeled locomotion and grasping, slope locomotion, block terrain locomotion, carrying a bag, and outdoor locomotion.
Disliked
jechoi@andrew.cmu.edu
A Mobile Quad-Arm Robot ARMS: Wheeled-Legged Tripedal Locomotion and Quad-Arm Loco-Manipulation : This article proposes a mobile quad-arm robot: ARMS, which unifies wheeled-legged tripedal locomotion, wheeled locomotion, and quad-arm loco-manipulation. ARMS's four arms have different mechanisms and are partially designed to be general-purpose arms for the hybrid locomotion and loco-manipulation. One three-degree-of-freedom (DOF) arm has an active wheel, which is used for wheeled-legged tripedal walking and wheeled driving with passive wheels attached to the torso. Two three-DOF general-purpose arms are series elastic and used for wheeled-legged tripedal walking, object grasping, and manipulation. The upper two-DOF arm is used for manipulation only; its position and orientation are determined by coordinating all arms. Each motor is controlled by an angle controller and trajectory modification with angle, angular velocity, angular acceleration, and torque constraints. ARMS was verified with seven experiments involving joint control, wheeled-legged locomotion, wheeled locomotion and grasping, slope locomotion, block terrain locomotion, carrying a bag, and outdoor locomotion.
0
jechoi@andrew.cmu.edu [SEP] A Mobile Quad-Arm Robot ARMS: Wheeled-Legged Tripedal Locomotion and Quad-Arm Loco-Manipulation : This article proposes a mobile quad-arm robot: ARMS, which unifies wheeled-legged tripedal locomotion, wheeled locomotion, and quad-arm loco-manipulation. ARMS's four arms have different mechanisms and are partially designed to be general-purpose arms for the hybrid locomotion and loco-manipulation. One three-degree-of-freedom (DOF) arm has an active wheel, which is used for wheeled-legged tripedal walking and wheeled driving with passive wheels attached to the torso. Two three-DOF general-purpose arms are series elastic and used for wheeled-legged tripedal walking, object grasping, and manipulation. The upper two-DOF arm is used for manipulation only; its position and orientation are determined by coordinating all arms. Each motor is controlled by an angle controller and trajectory modification with angle, angular velocity, angular acceleration, and torque constraints. ARMS was verified with seven experiments involving joint control, wheeled-legged locomotion, wheeled locomotion and grasping, slope locomotion, block terrain locomotion, carrying a bag, and outdoor locomotion.
15
Highly dynamic locomotion control of biped robot enhanced by swing arms
Swing arms have an irreplaceable role in promoting highly dynamic locomotion on bipedal robots by a larger angular momentum control space from the viewpoint of biomechanics. Few bipedal robots utilize swing arms and its redundancy characteristic of multiple degrees of freedom due to the lack of appropriate locomotion control strategies to perfectly integrate modeling and control. This paper presents a kind of control strategy by modeling the bipedal robot as a flywheel-spring loaded inverted pendulum (F-SLIP) to extract characteristics of swing arms and using the whole-body controller (WBC) to achieve these characteristics, and also proposes a evaluation system including three aspects of agility defined by us, stability and energy consumption for the highly dynamic locomotion of bipedal robots. We design several sets of simulation experiments and analyze the effects of swing arms according to the evaluation system during the jumping motion of Purple (Purple energy rises in the east)V1.0, a kind of bipedal robot designed to test high explosive locomotion. Results show that Purple's agility is increased by more than 10 percent, stabilization time is reduced by a factor of two, and energy consumption is reduced by more than 20 percent after introducing swing arms.
Disliked
jechoi@andrew.cmu.edu
Highly dynamic locomotion control of biped robot enhanced by swing arms : Swing arms have an irreplaceable role in promoting highly dynamic locomotion on bipedal robots by a larger angular momentum control space from the viewpoint of biomechanics. Few bipedal robots utilize swing arms and its redundancy characteristic of multiple degrees of freedom due to the lack of appropriate locomotion control strategies to perfectly integrate modeling and control. This paper presents a kind of control strategy by modeling the bipedal robot as a flywheel-spring loaded inverted pendulum (F-SLIP) to extract characteristics of swing arms and using the whole-body controller (WBC) to achieve these characteristics, and also proposes a evaluation system including three aspects of agility defined by us, stability and energy consumption for the highly dynamic locomotion of bipedal robots. We design several sets of simulation experiments and analyze the effects of swing arms according to the evaluation system during the jumping motion of Purple (Purple energy rises in the east)V1.0, a kind of bipedal robot designed to test high explosive locomotion. Results show that Purple's agility is increased by more than 10 percent, stabilization time is reduced by a factor of two, and energy consumption is reduced by more than 20 percent after introducing swing arms.
0
jechoi@andrew.cmu.edu [SEP] Highly dynamic locomotion control of biped robot enhanced by swing arms : Swing arms have an irreplaceable role in promoting highly dynamic locomotion on bipedal robots by a larger angular momentum control space from the viewpoint of biomechanics. Few bipedal robots utilize swing arms and its redundancy characteristic of multiple degrees of freedom due to the lack of appropriate locomotion control strategies to perfectly integrate modeling and control. This paper presents a kind of control strategy by modeling the bipedal robot as a flywheel-spring loaded inverted pendulum (F-SLIP) to extract characteristics of swing arms and using the whole-body controller (WBC) to achieve these characteristics, and also proposes a evaluation system including three aspects of agility defined by us, stability and energy consumption for the highly dynamic locomotion of bipedal robots. We design several sets of simulation experiments and analyze the effects of swing arms according to the evaluation system during the jumping motion of Purple (Purple energy rises in the east)V1.0, a kind of bipedal robot designed to test high explosive locomotion. Results show that Purple's agility is increased by more than 10 percent, stabilization time is reduced by a factor of two, and energy consumption is reduced by more than 20 percent after introducing swing arms.
425
Geometrization of deep networks for the interpretability of deep learning systems
How to understand deep learning systems remains an open problem. In this paper we propose that the answer may lie in the geometrization of deep networks. Geometrization is a bridge to connect physics, geometry, deep network and quantum computation and this may result in a new scheme to reveal the rule of the physical world. By comparing the geometry of image matching and deep networks, we show that geometrization of deep networks can be used to understand existing deep learning systems and it may also help to solve the interpretability problem of deep learning systems.
Disliked
zrz@andrew.cmu.edu
Geometrization of deep networks for the interpretability of deep learning systems : How to understand deep learning systems remains an open problem. In this paper we propose that the answer may lie in the geometrization of deep networks. Geometrization is a bridge to connect physics, geometry, deep network and quantum computation and this may result in a new scheme to reveal the rule of the physical world. By comparing the geometry of image matching and deep networks, we show that geometrization of deep networks can be used to understand existing deep learning systems and it may also help to solve the interpretability problem of deep learning systems.
0
zrz@andrew.cmu.edu [SEP] Geometrization of deep networks for the interpretability of deep learning systems : How to understand deep learning systems remains an open problem. In this paper we propose that the answer may lie in the geometrization of deep networks. Geometrization is a bridge to connect physics, geometry, deep network and quantum computation and this may result in a new scheme to reveal the rule of the physical world. By comparing the geometry of image matching and deep networks, we show that geometrization of deep networks can be used to understand existing deep learning systems and it may also help to solve the interpretability problem of deep learning systems.
160
Position Paper: Towards Transparent Machine Learning
Transparent machine learning is introduced as an alternative form of machine learning, where both the model and the learning system are represented in source code form. The goal of this project is to enable direct human understanding of machine learning models, giving us the ability to learn, verify, and refine them as programs. If solved, this technology could represent a best-case scenario for the safety and security of AI systems going forward.
Liked
zrz@andrew.cmu.edu
Position Paper: Towards Transparent Machine Learning : Transparent machine learning is introduced as an alternative form of machine learning, where both the model and the learning system are represented in source code form. The goal of this project is to enable direct human understanding of machine learning models, giving us the ability to learn, verify, and refine them as programs. If solved, this technology could represent a best-case scenario for the safety and security of AI systems going forward.
1
zrz@andrew.cmu.edu [SEP] Position Paper: Towards Transparent Machine Learning : Transparent machine learning is introduced as an alternative form of machine learning, where both the model and the learning system are represented in source code form. The goal of this project is to enable direct human understanding of machine learning models, giving us the ability to learn, verify, and refine them as programs. If solved, this technology could represent a best-case scenario for the safety and security of AI systems going forward.
31
Deep Learning: From Basics to Building Deep Neural Networks with Python
This book is intended for beginners who have no familiarity with deep learning. Our only expectation from readers is that they already have the basic programming skills in Python.
Disliked
zrz@andrew.cmu.edu
Deep Learning: From Basics to Building Deep Neural Networks with Python : This book is intended for beginners who have no familiarity with deep learning. Our only expectation from readers is that they already have the basic programming skills in Python.
0
zrz@andrew.cmu.edu [SEP] Deep Learning: From Basics to Building Deep Neural Networks with Python : This book is intended for beginners who have no familiarity with deep learning. Our only expectation from readers is that they already have the basic programming skills in Python.
202
Automating the Learning of Inverse Kinematics for Robotic Arms with Redundant DoFs
Inverse Kinematics (IK) solves the problem of mapping from the Cartesian space to the joint configuration space of a robotic arm. It has a wide range of applications in areas such as computer graphics, protein structure prediction, and robotics. With the vast advances of artificial neural networks (NNs), many researchers recently turned to data-driven approaches to solving the IK problem. Unfortunately, NNs become inadequate for robotic arms with redundant Degrees-of-Freedom (DoFs). This is because such arms may have multiple angle solutions to reach the same desired pose, while typical NNs only implement one-to-one mapping functions, which associate just one consistent output for a given input. In order to train usable NNs to solve the IK problem, most existing works employ customized training datasets, in which every desired pose only has one angle solution. This inevitably limits the generalization and automation of the proposed approaches. This paper breaks through at two fronts: (1) a systematic and mechanical approach to training data collection that covers the entire working space of the robotic arm, and can be fully automated and done only once after the arm is developed; and (2) a novel NN-based framework that can leverage the redundant DoFs to produce multiple angle solutions to any given desired pose of the robotic arm. The latter is especially useful for robotic applications such as obstacle avoidance and posture imitation.
Liked
jechoi@andrew.cmu.edu
Automating the Learning of Inverse Kinematics for Robotic Arms with Redundant DoFs : Inverse Kinematics (IK) solves the problem of mapping from the Cartesian space to the joint configuration space of a robotic arm. It has a wide range of applications in areas such as computer graphics, protein structure prediction, and robotics. With the vast advances of artificial neural networks (NNs), many researchers recently turned to data-driven approaches to solving the IK problem. Unfortunately, NNs become inadequate for robotic arms with redundant Degrees-of-Freedom (DoFs). This is because such arms may have multiple angle solutions to reach the same desired pose, while typical NNs only implement one-to-one mapping functions, which associate just one consistent output for a given input. In order to train usable NNs to solve the IK problem, most existing works employ customized training datasets, in which every desired pose only has one angle solution. This inevitably limits the generalization and automation of the proposed approaches. This paper breaks through at two fronts: (1) a systematic and mechanical approach to training data collection that covers the entire working space of the robotic arm, and can be fully automated and done only once after the arm is developed; and (2) a novel NN-based framework that can leverage the redundant DoFs to produce multiple angle solutions to any given desired pose of the robotic arm. The latter is especially useful for robotic applications such as obstacle avoidance and posture imitation.
1
jechoi@andrew.cmu.edu [SEP] Automating the Learning of Inverse Kinematics for Robotic Arms with Redundant DoFs : Inverse Kinematics (IK) solves the problem of mapping from the Cartesian space to the joint configuration space of a robotic arm. It has a wide range of applications in areas such as computer graphics, protein structure prediction, and robotics. With the vast advances of artificial neural networks (NNs), many researchers recently turned to data-driven approaches to solving the IK problem. Unfortunately, NNs become inadequate for robotic arms with redundant Degrees-of-Freedom (DoFs). This is because such arms may have multiple angle solutions to reach the same desired pose, while typical NNs only implement one-to-one mapping functions, which associate just one consistent output for a given input. In order to train usable NNs to solve the IK problem, most existing works employ customized training datasets, in which every desired pose only has one angle solution. This inevitably limits the generalization and automation of the proposed approaches. This paper breaks through at two fronts: (1) a systematic and mechanical approach to training data collection that covers the entire working space of the robotic arm, and can be fully automated and done only once after the arm is developed; and (2) a novel NN-based framework that can leverage the redundant DoFs to produce multiple angle solutions to any given desired pose of the robotic arm. The latter is especially useful for robotic applications such as obstacle avoidance and posture imitation.
455
Classic machine learning methods
In this chapter, we present the main classic machine learning methods. A large part of the chapter is devoted to supervised learning techniques for classification and regression, including nearest-neighbor methods, linear and logistic regressions, support vector machines and tree-based algorithms. We also describe the problem of overfitting as well as strategies to overcome it. We finally provide a brief overview of unsupervised learning methods, namely for clustering and dimensionality reduction.
Disliked
zrz@andrew.cmu.edu
Classic machine learning methods : In this chapter, we present the main classic machine learning methods. A large part of the chapter is devoted to supervised learning techniques for classification and regression, including nearest-neighbor methods, linear and logistic regressions, support vector machines and tree-based algorithms. We also describe the problem of overfitting as well as strategies to overcome it. We finally provide a brief overview of unsupervised learning methods, namely for clustering and dimensionality reduction.
0
zrz@andrew.cmu.edu [SEP] Classic machine learning methods : In this chapter, we present the main classic machine learning methods. A large part of the chapter is devoted to supervised learning techniques for classification and regression, including nearest-neighbor methods, linear and logistic regressions, support vector machines and tree-based algorithms. We also describe the problem of overfitting as well as strategies to overcome it. We finally provide a brief overview of unsupervised learning methods, namely for clustering and dimensionality reduction.
123
Rope3D: TheRoadside Perception Dataset for Autonomous Driving and Monocular 3D Object Detection Task
Concurrent perception datasets for autonomous driving are mainly limited to frontal view with sensors mounted on the vehicle. None of them is designed for the overlooked roadside perception tasks. On the other hand, the data captured from roadside cameras have strengths over frontal-view data, which is believed to facilitate a safer and more intelligent autonomous driving system. To accelerate the progress of roadside perception, we present the first high-diversity challenging Roadside Perception 3D dataset- Rope3D from a novel view. The dataset consists of 50k images and over 1.5M 3D objects in various scenes, which are captured under different settings including various cameras with ambiguous mounting positions, camera specifications, viewpoints, and different environmental conditions. We conduct strict 2D-3D joint annotation and comprehensive data analysis, as well as set up a new 3D roadside perception benchmark with metrics and evaluation devkit. Furthermore, we tailor the existing frontal-view monocular 3D object detection approaches and propose to leverage the geometry constraint to solve the inherent ambiguities caused by various sensors, viewpoints. Our dataset is available on https://thudair.baai.ac.cn/rope.
Disliked
zrz@andrew.cmu.edu
Rope3D: TheRoadside Perception Dataset for Autonomous Driving and Monocular 3D Object Detection Task : Concurrent perception datasets for autonomous driving are mainly limited to frontal view with sensors mounted on the vehicle. None of them is designed for the overlooked roadside perception tasks. On the other hand, the data captured from roadside cameras have strengths over frontal-view data, which is believed to facilitate a safer and more intelligent autonomous driving system. To accelerate the progress of roadside perception, we present the first high-diversity challenging Roadside Perception 3D dataset- Rope3D from a novel view. The dataset consists of 50k images and over 1.5M 3D objects in various scenes, which are captured under different settings including various cameras with ambiguous mounting positions, camera specifications, viewpoints, and different environmental conditions. We conduct strict 2D-3D joint annotation and comprehensive data analysis, as well as set up a new 3D roadside perception benchmark with metrics and evaluation devkit. Furthermore, we tailor the existing frontal-view monocular 3D object detection approaches and propose to leverage the geometry constraint to solve the inherent ambiguities caused by various sensors, viewpoints. Our dataset is available on https://thudair.baai.ac.cn/rope.
0
zrz@andrew.cmu.edu [SEP] Rope3D: TheRoadside Perception Dataset for Autonomous Driving and Monocular 3D Object Detection Task : Concurrent perception datasets for autonomous driving are mainly limited to frontal view with sensors mounted on the vehicle. None of them is designed for the overlooked roadside perception tasks. On the other hand, the data captured from roadside cameras have strengths over frontal-view data, which is believed to facilitate a safer and more intelligent autonomous driving system. To accelerate the progress of roadside perception, we present the first high-diversity challenging Roadside Perception 3D dataset- Rope3D from a novel view. The dataset consists of 50k images and over 1.5M 3D objects in various scenes, which are captured under different settings including various cameras with ambiguous mounting positions, camera specifications, viewpoints, and different environmental conditions. We conduct strict 2D-3D joint annotation and comprehensive data analysis, as well as set up a new 3D roadside perception benchmark with metrics and evaluation devkit. Furthermore, we tailor the existing frontal-view monocular 3D object detection approaches and propose to leverage the geometry constraint to solve the inherent ambiguities caused by various sensors, viewpoints. Our dataset is available on https://thudair.baai.ac.cn/rope.
304
A metric for characterizing the arm nonuse workspace in poststroke individuals using a robot arm
An over-reliance on the less-affected limb for functional tasks at the expense of the paretic limb and in spite of recovered capacity is an often-observed phenomenon in survivors of hemispheric stroke. The difference between capacity for use and actual spontaneous use is referred to as arm nonuse. Obtaining an ecologically valid evaluation of arm nonuse is challenging because it requires the observation of spontaneous arm choice for different tasks, which can easily be influenced by instructions, presumed expectations, and awareness that one is being tested. To better quantify arm nonuse, we developed the Bimanual Arm Reaching Test with a Robot (BARTR) for quantitatively assessing arm nonuse in chronic stroke survivors. The BARTR is an instrument that utilizes a robot arm as a means of remote and unbiased data collection of nuanced spatial data for clinical evaluations of arm nonuse. This approach shows promise for determining the efficacy of interventions designed to reduce paretic arm nonuse and enhance functional recovery after stroke. We show that the BARTR satisfies the criteria of an appropriate metric for neurorehabilitative contexts: it is valid, reliable, and simple to use.
Liked
jechoi@andrew.cmu.edu
A metric for characterizing the arm nonuse workspace in poststroke individuals using a robot arm : An over-reliance on the less-affected limb for functional tasks at the expense of the paretic limb and in spite of recovered capacity is an often-observed phenomenon in survivors of hemispheric stroke. The difference between capacity for use and actual spontaneous use is referred to as arm nonuse. Obtaining an ecologically valid evaluation of arm nonuse is challenging because it requires the observation of spontaneous arm choice for different tasks, which can easily be influenced by instructions, presumed expectations, and awareness that one is being tested. To better quantify arm nonuse, we developed the Bimanual Arm Reaching Test with a Robot (BARTR) for quantitatively assessing arm nonuse in chronic stroke survivors. The BARTR is an instrument that utilizes a robot arm as a means of remote and unbiased data collection of nuanced spatial data for clinical evaluations of arm nonuse. This approach shows promise for determining the efficacy of interventions designed to reduce paretic arm nonuse and enhance functional recovery after stroke. We show that the BARTR satisfies the criteria of an appropriate metric for neurorehabilitative contexts: it is valid, reliable, and simple to use.
1
jechoi@andrew.cmu.edu [SEP] A metric for characterizing the arm nonuse workspace in poststroke individuals using a robot arm : An over-reliance on the less-affected limb for functional tasks at the expense of the paretic limb and in spite of recovered capacity is an often-observed phenomenon in survivors of hemispheric stroke. The difference between capacity for use and actual spontaneous use is referred to as arm nonuse. Obtaining an ecologically valid evaluation of arm nonuse is challenging because it requires the observation of spontaneous arm choice for different tasks, which can easily be influenced by instructions, presumed expectations, and awareness that one is being tested. To better quantify arm nonuse, we developed the Bimanual Arm Reaching Test with a Robot (BARTR) for quantitatively assessing arm nonuse in chronic stroke survivors. The BARTR is an instrument that utilizes a robot arm as a means of remote and unbiased data collection of nuanced spatial data for clinical evaluations of arm nonuse. This approach shows promise for determining the efficacy of interventions designed to reduce paretic arm nonuse and enhance functional recovery after stroke. We show that the BARTR satisfies the criteria of an appropriate metric for neurorehabilitative contexts: it is valid, reliable, and simple to use.
410
Probabilistic Deep Learning with Probabilistic Neural Networks and Deep Probabilistic Models
Probabilistic deep learning is deep learning that accounts for uncertainty, both model uncertainty and data uncertainty. It is based on the use of probabilistic models and deep neural networks. We distinguish two approaches to probabilistic deep learning: probabilistic neural networks and deep probabilistic models. The former employs deep neural networks that utilize probabilistic layers which can represent and process uncertainty; the latter uses probabilistic models that incorporate deep neural network components which capture complex non-linear stochastic relationships between the random variables. We discuss some major examples of each approach including Bayesian neural networks and mixture density networks (for probabilistic neural networks), and variational autoencoders, deep Gaussian processes and deep mixed effects models (for deep probabilistic models). TensorFlow Probability is a library for probabilistic modeling and inference which can be used for both approaches of probabilistic deep learning. We include its code examples for illustration.
Liked
zrz@andrew.cmu.edu
Probabilistic Deep Learning with Probabilistic Neural Networks and Deep Probabilistic Models : Probabilistic deep learning is deep learning that accounts for uncertainty, both model uncertainty and data uncertainty. It is based on the use of probabilistic models and deep neural networks. We distinguish two approaches to probabilistic deep learning: probabilistic neural networks and deep probabilistic models. The former employs deep neural networks that utilize probabilistic layers which can represent and process uncertainty; the latter uses probabilistic models that incorporate deep neural network components which capture complex non-linear stochastic relationships between the random variables. We discuss some major examples of each approach including Bayesian neural networks and mixture density networks (for probabilistic neural networks), and variational autoencoders, deep Gaussian processes and deep mixed effects models (for deep probabilistic models). TensorFlow Probability is a library for probabilistic modeling and inference which can be used for both approaches of probabilistic deep learning. We include its code examples for illustration.
1
zrz@andrew.cmu.edu [SEP] Probabilistic Deep Learning with Probabilistic Neural Networks and Deep Probabilistic Models : Probabilistic deep learning is deep learning that accounts for uncertainty, both model uncertainty and data uncertainty. It is based on the use of probabilistic models and deep neural networks. We distinguish two approaches to probabilistic deep learning: probabilistic neural networks and deep probabilistic models. The former employs deep neural networks that utilize probabilistic layers which can represent and process uncertainty; the latter uses probabilistic models that incorporate deep neural network components which capture complex non-linear stochastic relationships between the random variables. We discuss some major examples of each approach including Bayesian neural networks and mixture density networks (for probabilistic neural networks), and variational autoencoders, deep Gaussian processes and deep mixed effects models (for deep probabilistic models). TensorFlow Probability is a library for probabilistic modeling and inference which can be used for both approaches of probabilistic deep learning. We include its code examples for illustration.
169
A Survey on Resilient Machine Learning
Machine learning based system are increasingly being used for sensitive tasks such as security surveillance, guiding autonomous vehicle, taking investment decisions, detecting and blocking network intrusion and malware etc. However, recent research has shown that machine learning models are venerable to attacks by adversaries at all phases of machine learning (eg, training data collection, training, operation). All model classes of machine learning systems can be misled by providing carefully crafted inputs making them wrongly classify inputs. Maliciously created input samples can affect the learning process of a ML system by either slowing down the learning process, or affecting the performance of the learned mode, or causing the system make error(s) only in attacker's planned scenario. Because of these developments, understanding security of machine learning algorithms and systems is emerging as an important research area among computer security and machine learning researchers and practitioners. We present a survey of this emerging area in machine learning.
Liked
zrz@andrew.cmu.edu
A Survey on Resilient Machine Learning : Machine learning based system are increasingly being used for sensitive tasks such as security surveillance, guiding autonomous vehicle, taking investment decisions, detecting and blocking network intrusion and malware etc. However, recent research has shown that machine learning models are venerable to attacks by adversaries at all phases of machine learning (eg, training data collection, training, operation). All model classes of machine learning systems can be misled by providing carefully crafted inputs making them wrongly classify inputs. Maliciously created input samples can affect the learning process of a ML system by either slowing down the learning process, or affecting the performance of the learned mode, or causing the system make error(s) only in attacker's planned scenario. Because of these developments, understanding security of machine learning algorithms and systems is emerging as an important research area among computer security and machine learning researchers and practitioners. We present a survey of this emerging area in machine learning.
1
zrz@andrew.cmu.edu [SEP] A Survey on Resilient Machine Learning : Machine learning based system are increasingly being used for sensitive tasks such as security surveillance, guiding autonomous vehicle, taking investment decisions, detecting and blocking network intrusion and malware etc. However, recent research has shown that machine learning models are venerable to attacks by adversaries at all phases of machine learning (eg, training data collection, training, operation). All model classes of machine learning systems can be misled by providing carefully crafted inputs making them wrongly classify inputs. Maliciously created input samples can affect the learning process of a ML system by either slowing down the learning process, or affecting the performance of the learned mode, or causing the system make error(s) only in attacker's planned scenario. Because of these developments, understanding security of machine learning algorithms and systems is emerging as an important research area among computer security and machine learning researchers and practitioners. We present a survey of this emerging area in machine learning.
96
Security of Deep Learning Methodologies: Challenges and Opportunities
Despite the plethora of studies about security vulnerabilities and defenses of deep learning models, security aspects of deep learning methodologies, such as transfer learning, have been rarely studied. In this article, we highlight the security challenges and research opportunities of these methodologies, focusing on vulnerabilities and attacks unique to them.
Liked
zrz@andrew.cmu.edu
Security of Deep Learning Methodologies: Challenges and Opportunities : Despite the plethora of studies about security vulnerabilities and defenses of deep learning models, security aspects of deep learning methodologies, such as transfer learning, have been rarely studied. In this article, we highlight the security challenges and research opportunities of these methodologies, focusing on vulnerabilities and attacks unique to them.
1
zrz@andrew.cmu.edu [SEP] Security of Deep Learning Methodologies: Challenges and Opportunities : Despite the plethora of studies about security vulnerabilities and defenses of deep learning models, security aspects of deep learning methodologies, such as transfer learning, have been rarely studied. In this article, we highlight the security challenges and research opportunities of these methodologies, focusing on vulnerabilities and attacks unique to them.
265
Expanding the Boundaries of Vision Prior Knowledge in Multi-modal Large Language Models
Does the prior knowledge of the vision encoder constrain the capability boundary of Multi-modal Large Language Models (MLLMs)? While most existing research treats MLLMs as unified systems optimized through end-to-end training, the impact of vision encoder's prior knowledge is seldom investigated. In this work, we introduce a novel metric, $Rank_e$, to quantify the effect of prior knowledge of the vision encoder on MLLM performance. Our analysis reveals a positive correlation between prior knowledge and MLLM performance. Moreover, we find that domain-specific fine-tuning using solely end-to-end visual question answering (VQA) data is insufficient, particularly for entities with low inherent visual prior knowledge. To address this issue, we propose VisPRE (Vision Prior Remediation), a two-stage training framework that explicitly incorporates prior knowledge at the vision encoder level. Experimental results demonstrate that augmenting vision encoder's prior knowledge substantially boosts the visual understanding capabilities of MLLMs, offering a novel and effective strategy for improving performance, especially in scenarios involving uncommon visual entities.
Liked
zrz@andrew.cmu.edu
Expanding the Boundaries of Vision Prior Knowledge in Multi-modal Large Language Models : Does the prior knowledge of the vision encoder constrain the capability boundary of Multi-modal Large Language Models (MLLMs)? While most existing research treats MLLMs as unified systems optimized through end-to-end training, the impact of vision encoder's prior knowledge is seldom investigated. In this work, we introduce a novel metric, $Rank_e$, to quantify the effect of prior knowledge of the vision encoder on MLLM performance. Our analysis reveals a positive correlation between prior knowledge and MLLM performance. Moreover, we find that domain-specific fine-tuning using solely end-to-end visual question answering (VQA) data is insufficient, particularly for entities with low inherent visual prior knowledge. To address this issue, we propose VisPRE (Vision Prior Remediation), a two-stage training framework that explicitly incorporates prior knowledge at the vision encoder level. Experimental results demonstrate that augmenting vision encoder's prior knowledge substantially boosts the visual understanding capabilities of MLLMs, offering a novel and effective strategy for improving performance, especially in scenarios involving uncommon visual entities.
1
zrz@andrew.cmu.edu [SEP] Expanding the Boundaries of Vision Prior Knowledge in Multi-modal Large Language Models : Does the prior knowledge of the vision encoder constrain the capability boundary of Multi-modal Large Language Models (MLLMs)? While most existing research treats MLLMs as unified systems optimized through end-to-end training, the impact of vision encoder's prior knowledge is seldom investigated. In this work, we introduce a novel metric, $Rank_e$, to quantify the effect of prior knowledge of the vision encoder on MLLM performance. Our analysis reveals a positive correlation between prior knowledge and MLLM performance. Moreover, we find that domain-specific fine-tuning using solely end-to-end visual question answering (VQA) data is insufficient, particularly for entities with low inherent visual prior knowledge. To address this issue, we propose VisPRE (Vision Prior Remediation), a two-stage training framework that explicitly incorporates prior knowledge at the vision encoder level. Experimental results demonstrate that augmenting vision encoder's prior knowledge substantially boosts the visual understanding capabilities of MLLMs, offering a novel and effective strategy for improving performance, especially in scenarios involving uncommon visual entities.
348
Neural Models and Algorithms for Sensorimotor Control of an Octopus Arm
In this article, a biophysically realistic model of a soft octopus arm with internal musculature is presented. The modeling is motivated by experimental observations of sensorimotor control where an arm localizes and reaches a target. Major contributions of this article are: (i) development of models to capture the mechanical properties of arm musculature, the electrical properties of the arm peripheral nervous system (PNS), and the coupling of PNS with muscular contractions; (ii) modeling the arm sensory system, including chemosensing and proprioception; and (iii) algorithms for sensorimotor control, which include a novel feedback neural motor control law for mimicking target-oriented arm reaching motions, and a novel consensus algorithm for solving sensing problems such as locating a food source from local chemical sensory information (exogenous) and arm deformation information (endogenous). Several analytical results, including rest-state characterization and stability properties of the proposed sensing and motor control algorithms, are provided. Numerical simulations demonstrate the efficacy of our approach. Qualitative comparisons against observed arm rest shapes and target-oriented reaching motions are also reported.
Liked
jechoi@andrew.cmu.edu
Neural Models and Algorithms for Sensorimotor Control of an Octopus Arm : In this article, a biophysically realistic model of a soft octopus arm with internal musculature is presented. The modeling is motivated by experimental observations of sensorimotor control where an arm localizes and reaches a target. Major contributions of this article are: (i) development of models to capture the mechanical properties of arm musculature, the electrical properties of the arm peripheral nervous system (PNS), and the coupling of PNS with muscular contractions; (ii) modeling the arm sensory system, including chemosensing and proprioception; and (iii) algorithms for sensorimotor control, which include a novel feedback neural motor control law for mimicking target-oriented arm reaching motions, and a novel consensus algorithm for solving sensing problems such as locating a food source from local chemical sensory information (exogenous) and arm deformation information (endogenous). Several analytical results, including rest-state characterization and stability properties of the proposed sensing and motor control algorithms, are provided. Numerical simulations demonstrate the efficacy of our approach. Qualitative comparisons against observed arm rest shapes and target-oriented reaching motions are also reported.
1
jechoi@andrew.cmu.edu [SEP] Neural Models and Algorithms for Sensorimotor Control of an Octopus Arm : In this article, a biophysically realistic model of a soft octopus arm with internal musculature is presented. The modeling is motivated by experimental observations of sensorimotor control where an arm localizes and reaches a target. Major contributions of this article are: (i) development of models to capture the mechanical properties of arm musculature, the electrical properties of the arm peripheral nervous system (PNS), and the coupling of PNS with muscular contractions; (ii) modeling the arm sensory system, including chemosensing and proprioception; and (iii) algorithms for sensorimotor control, which include a novel feedback neural motor control law for mimicking target-oriented arm reaching motions, and a novel consensus algorithm for solving sensing problems such as locating a food source from local chemical sensory information (exogenous) and arm deformation information (endogenous). Several analytical results, including rest-state characterization and stability properties of the proposed sensing and motor control algorithms, are provided. Numerical simulations demonstrate the efficacy of our approach. Qualitative comparisons against observed arm rest shapes and target-oriented reaching motions are also reported.
493
PyKale: Knowledge-Aware Machine Learning from Multiple Sources in Python
Machine learning is a general-purpose technology holding promises for many interdisciplinary research problems. However, significant barriers exist in crossing disciplinary boundaries when most machine learning tools are developed in different areas separately. We present Pykale - a Python library for knowledge-aware machine learning on graphs, images, texts, and videos to enable and accelerate interdisciplinary research. We formulate new green machine learning guidelines based on standard software engineering practices and propose a novel pipeline-based application programming interface (API). PyKale focuses on leveraging knowledge from multiple sources for accurate and interpretable prediction, thus supporting multimodal learning and transfer learning (particularly domain adaptation) with latest deep learning and dimensionality reduction models. We build PyKale on PyTorch and leverage the rich PyTorch ecosystem. Our pipeline-based API design enforces standardization and minimalism, embracing green machine learning concepts via reducing repetitions and redundancy, reusing existing resources, and recycling learning models across areas. We demonstrate its interdisciplinary nature via examples in bioinformatics, knowledge graph, image/video recognition, and medical imaging.
Disliked
zrz@andrew.cmu.edu
PyKale: Knowledge-Aware Machine Learning from Multiple Sources in Python : Machine learning is a general-purpose technology holding promises for many interdisciplinary research problems. However, significant barriers exist in crossing disciplinary boundaries when most machine learning tools are developed in different areas separately. We present Pykale - a Python library for knowledge-aware machine learning on graphs, images, texts, and videos to enable and accelerate interdisciplinary research. We formulate new green machine learning guidelines based on standard software engineering practices and propose a novel pipeline-based application programming interface (API). PyKale focuses on leveraging knowledge from multiple sources for accurate and interpretable prediction, thus supporting multimodal learning and transfer learning (particularly domain adaptation) with latest deep learning and dimensionality reduction models. We build PyKale on PyTorch and leverage the rich PyTorch ecosystem. Our pipeline-based API design enforces standardization and minimalism, embracing green machine learning concepts via reducing repetitions and redundancy, reusing existing resources, and recycling learning models across areas. We demonstrate its interdisciplinary nature via examples in bioinformatics, knowledge graph, image/video recognition, and medical imaging.
0
zrz@andrew.cmu.edu [SEP] PyKale: Knowledge-Aware Machine Learning from Multiple Sources in Python : Machine learning is a general-purpose technology holding promises for many interdisciplinary research problems. However, significant barriers exist in crossing disciplinary boundaries when most machine learning tools are developed in different areas separately. We present Pykale - a Python library for knowledge-aware machine learning on graphs, images, texts, and videos to enable and accelerate interdisciplinary research. We formulate new green machine learning guidelines based on standard software engineering practices and propose a novel pipeline-based application programming interface (API). PyKale focuses on leveraging knowledge from multiple sources for accurate and interpretable prediction, thus supporting multimodal learning and transfer learning (particularly domain adaptation) with latest deep learning and dimensionality reduction models. We build PyKale on PyTorch and leverage the rich PyTorch ecosystem. Our pipeline-based API design enforces standardization and minimalism, embracing green machine learning concepts via reducing repetitions and redundancy, reusing existing resources, and recycling learning models across areas. We demonstrate its interdisciplinary nature via examples in bioinformatics, knowledge graph, image/video recognition, and medical imaging.
136
Improving Cancer Imaging Diagnosis with Bayesian Networks and Deep Learning: A Bayesian Deep Learning Approach
With recent advancements in the development of artificial intelligence applications using theories and algorithms in machine learning, many accurate models can be created to train and predict on given datasets. With the realization of the importance of imaging interpretation in cancer diagnosis, this article aims to investigate the theory behind Deep Learning and Bayesian Network prediction models. Based on the advantages and drawbacks of each model, different approaches will be used to construct a Bayesian Deep Learning Model, combining the strengths while minimizing the weaknesses. Finally, the applications and accuracy of the resulting Bayesian Deep Learning approach in the health industry in classifying images will be analyzed.
Liked
zrz@andrew.cmu.edu
Improving Cancer Imaging Diagnosis with Bayesian Networks and Deep Learning: A Bayesian Deep Learning Approach : With recent advancements in the development of artificial intelligence applications using theories and algorithms in machine learning, many accurate models can be created to train and predict on given datasets. With the realization of the importance of imaging interpretation in cancer diagnosis, this article aims to investigate the theory behind Deep Learning and Bayesian Network prediction models. Based on the advantages and drawbacks of each model, different approaches will be used to construct a Bayesian Deep Learning Model, combining the strengths while minimizing the weaknesses. Finally, the applications and accuracy of the resulting Bayesian Deep Learning approach in the health industry in classifying images will be analyzed.
1
zrz@andrew.cmu.edu [SEP] Improving Cancer Imaging Diagnosis with Bayesian Networks and Deep Learning: A Bayesian Deep Learning Approach : With recent advancements in the development of artificial intelligence applications using theories and algorithms in machine learning, many accurate models can be created to train and predict on given datasets. With the realization of the importance of imaging interpretation in cancer diagnosis, this article aims to investigate the theory behind Deep Learning and Bayesian Network prediction models. Based on the advantages and drawbacks of each model, different approaches will be used to construct a Bayesian Deep Learning Model, combining the strengths while minimizing the weaknesses. Finally, the applications and accuracy of the resulting Bayesian Deep Learning approach in the health industry in classifying images will be analyzed.
217
Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning
Deep reinforcement learning (deep RL) has achieved superior performance in complex sequential tasks by using deep neural networks as function approximators to learn directly from raw input images. However, learning directly from raw images is data inefficient. The agent must learn feature representation of complex states in addition to learning a policy. As a result, deep RL typically suffers from slow learning speeds and often requires a prohibitively large amount of training time and data to reach reasonable performance, making it inapplicable to real-world settings where data is expensive. In this work, we improve data efficiency in deep RL by addressing one of the two learning goals, feature learning. We leverage supervised learning to pre-train on a small set of non-expert human demonstrations and empirically evaluate our approach using the asynchronous advantage actor-critic algorithms (A3C) in the Atari domain. Our results show significant improvements in learning speed, even when the provided demonstration is noisy and of low quality.
Liked
zrz@andrew.cmu.edu
Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning : Deep reinforcement learning (deep RL) has achieved superior performance in complex sequential tasks by using deep neural networks as function approximators to learn directly from raw input images. However, learning directly from raw images is data inefficient. The agent must learn feature representation of complex states in addition to learning a policy. As a result, deep RL typically suffers from slow learning speeds and often requires a prohibitively large amount of training time and data to reach reasonable performance, making it inapplicable to real-world settings where data is expensive. In this work, we improve data efficiency in deep RL by addressing one of the two learning goals, feature learning. We leverage supervised learning to pre-train on a small set of non-expert human demonstrations and empirically evaluate our approach using the asynchronous advantage actor-critic algorithms (A3C) in the Atari domain. Our results show significant improvements in learning speed, even when the provided demonstration is noisy and of low quality.
1
zrz@andrew.cmu.edu [SEP] Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning : Deep reinforcement learning (deep RL) has achieved superior performance in complex sequential tasks by using deep neural networks as function approximators to learn directly from raw input images. However, learning directly from raw images is data inefficient. The agent must learn feature representation of complex states in addition to learning a policy. As a result, deep RL typically suffers from slow learning speeds and often requires a prohibitively large amount of training time and data to reach reasonable performance, making it inapplicable to real-world settings where data is expensive. In this work, we improve data efficiency in deep RL by addressing one of the two learning goals, feature learning. We leverage supervised learning to pre-train on a small set of non-expert human demonstrations and empirically evaluate our approach using the asynchronous advantage actor-critic algorithms (A3C) in the Atari domain. Our results show significant improvements in learning speed, even when the provided demonstration is noisy and of low quality.
266
Generating and Customizing Robotic Arm Trajectories using Neural Networks
We introduce a neural network approach for generating and customizing the trajectory of a robotic arm, that guarantees precision and repeatability. To highlight the potential of this novel method, we describe the design and implementation of the technique and show its application in an experimental setting of cognitive robotics. In this scenario, the NICO robot was characterized by the ability to point to specific points in space with precise linear movements, increasing the predictability of the robotic action during its interaction with humans. To achieve this goal, the neural network computes the forward kinematics of the robot arm. By integrating it with a generator of joint angles, another neural network was developed and trained on an artificial dataset created from suitable start and end poses of the robotic arm. Through the computation of angular velocities, the robot was characterized by its ability to perform the movement, and the quality of its action was evaluated in terms of shape and accuracy. Thanks to its broad applicability, our approach successfully generates precise trajectories that could be customized in their shape and adapted to different settings.
Liked
jechoi@andrew.cmu.edu
Generating and Customizing Robotic Arm Trajectories using Neural Networks : We introduce a neural network approach for generating and customizing the trajectory of a robotic arm, that guarantees precision and repeatability. To highlight the potential of this novel method, we describe the design and implementation of the technique and show its application in an experimental setting of cognitive robotics. In this scenario, the NICO robot was characterized by the ability to point to specific points in space with precise linear movements, increasing the predictability of the robotic action during its interaction with humans. To achieve this goal, the neural network computes the forward kinematics of the robot arm. By integrating it with a generator of joint angles, another neural network was developed and trained on an artificial dataset created from suitable start and end poses of the robotic arm. Through the computation of angular velocities, the robot was characterized by its ability to perform the movement, and the quality of its action was evaluated in terms of shape and accuracy. Thanks to its broad applicability, our approach successfully generates precise trajectories that could be customized in their shape and adapted to different settings.
1
jechoi@andrew.cmu.edu [SEP] Generating and Customizing Robotic Arm Trajectories using Neural Networks : We introduce a neural network approach for generating and customizing the trajectory of a robotic arm, that guarantees precision and repeatability. To highlight the potential of this novel method, we describe the design and implementation of the technique and show its application in an experimental setting of cognitive robotics. In this scenario, the NICO robot was characterized by the ability to point to specific points in space with precise linear movements, increasing the predictability of the robotic action during its interaction with humans. To achieve this goal, the neural network computes the forward kinematics of the robot arm. By integrating it with a generator of joint angles, another neural network was developed and trained on an artificial dataset created from suitable start and end poses of the robotic arm. Through the computation of angular velocities, the robot was characterized by its ability to perform the movement, and the quality of its action was evaluated in terms of shape and accuracy. Thanks to its broad applicability, our approach successfully generates precise trajectories that could be customized in their shape and adapted to different settings.
454
Pedipulate: Enabling Manipulation Skills using a Quadruped Robot's Leg
Legged robots have the potential to become vital in maintenance, home support, and exploration scenarios. In order to interact with and manipulate their environments, most legged robots are equipped with a dedicated robot arm, which means additional mass and mechanical complexity compared to standard legged robots. In this work, we explore pedipulation - using the legs of a legged robot for manipulation. By training a reinforcement learning policy that tracks position targets for one foot, we enable a dedicated pedipulation controller that is robust to disturbances, has a large workspace through whole-body behaviors, and can reach far-away targets with gait emergence, enabling loco-pedipulation. By deploying our controller on a quadrupedal robot using teleoperation, we demonstrate various real-world tasks such as door opening, sample collection, and pushing obstacles. We demonstrate load carrying of more than 2.0 kg at the foot. Additionally, the controller is robust to interaction forces at the foot, disturbances at the base, and slippery contact surfaces. Videos of the experiments are available at https://sites.google.com/leggedrobotics.com/pedipulate.
Liked
jechoi@andrew.cmu.edu
Pedipulate: Enabling Manipulation Skills using a Quadruped Robot's Leg : Legged robots have the potential to become vital in maintenance, home support, and exploration scenarios. In order to interact with and manipulate their environments, most legged robots are equipped with a dedicated robot arm, which means additional mass and mechanical complexity compared to standard legged robots. In this work, we explore pedipulation - using the legs of a legged robot for manipulation. By training a reinforcement learning policy that tracks position targets for one foot, we enable a dedicated pedipulation controller that is robust to disturbances, has a large workspace through whole-body behaviors, and can reach far-away targets with gait emergence, enabling loco-pedipulation. By deploying our controller on a quadrupedal robot using teleoperation, we demonstrate various real-world tasks such as door opening, sample collection, and pushing obstacles. We demonstrate load carrying of more than 2.0 kg at the foot. Additionally, the controller is robust to interaction forces at the foot, disturbances at the base, and slippery contact surfaces. Videos of the experiments are available at https://sites.google.com/leggedrobotics.com/pedipulate.
1
jechoi@andrew.cmu.edu [SEP] Pedipulate: Enabling Manipulation Skills using a Quadruped Robot's Leg : Legged robots have the potential to become vital in maintenance, home support, and exploration scenarios. In order to interact with and manipulate their environments, most legged robots are equipped with a dedicated robot arm, which means additional mass and mechanical complexity compared to standard legged robots. In this work, we explore pedipulation - using the legs of a legged robot for manipulation. By training a reinforcement learning policy that tracks position targets for one foot, we enable a dedicated pedipulation controller that is robust to disturbances, has a large workspace through whole-body behaviors, and can reach far-away targets with gait emergence, enabling loco-pedipulation. By deploying our controller on a quadrupedal robot using teleoperation, we demonstrate various real-world tasks such as door opening, sample collection, and pushing obstacles. We demonstrate load carrying of more than 2.0 kg at the foot. Additionally, the controller is robust to interaction forces at the foot, disturbances at the base, and slippery contact surfaces. Videos of the experiments are available at https://sites.google.com/leggedrobotics.com/pedipulate.
533
Augmented Reality Remote Operation of Dual Arm Manipulators in Hot Boxes
In nuclear isotope and chemistry laboratories, hot cells and gloveboxes provide scientists with a controlled and safe environment to perform experiments. Working on experiments in these isolated containment cells requires scientists to be physically present. For hot cell work today, scientists manipulate equipment and radioactive material inside through a bilateral mechanical control mechanism. Motions produced outside the cell with the master control levers are mechanically transferred to the internal grippers inside the shielded containment cell. There is a growing need to have the capability to conduct experiments within these cells remotely. A simple method to enable remote manipulations within hot cell and glovebox cells is to mount two robotic arms inside a box to mimic the motions of human hands. An AR application was built in this work to allow a user wearing a Microsoft HoloLens 2 headset to teleoperate dual arm manipulators by grasping robotic end-effector digital replicas in AR from a remote location. In addition to the real-time replica of the physical robotic arms in AR, the application enables users to view a live video stream attached to the robotic arms and parse a 3D point cloud of 3D objects in their remote AR environment for better situational awareness. This work also provides users with virtual fixture to assist in manipulation and other teleoperation tasks.
Liked
jechoi@andrew.cmu.edu
Augmented Reality Remote Operation of Dual Arm Manipulators in Hot Boxes : In nuclear isotope and chemistry laboratories, hot cells and gloveboxes provide scientists with a controlled and safe environment to perform experiments. Working on experiments in these isolated containment cells requires scientists to be physically present. For hot cell work today, scientists manipulate equipment and radioactive material inside through a bilateral mechanical control mechanism. Motions produced outside the cell with the master control levers are mechanically transferred to the internal grippers inside the shielded containment cell. There is a growing need to have the capability to conduct experiments within these cells remotely. A simple method to enable remote manipulations within hot cell and glovebox cells is to mount two robotic arms inside a box to mimic the motions of human hands. An AR application was built in this work to allow a user wearing a Microsoft HoloLens 2 headset to teleoperate dual arm manipulators by grasping robotic end-effector digital replicas in AR from a remote location. In addition to the real-time replica of the physical robotic arms in AR, the application enables users to view a live video stream attached to the robotic arms and parse a 3D point cloud of 3D objects in their remote AR environment for better situational awareness. This work also provides users with virtual fixture to assist in manipulation and other teleoperation tasks.
1
jechoi@andrew.cmu.edu [SEP] Augmented Reality Remote Operation of Dual Arm Manipulators in Hot Boxes : In nuclear isotope and chemistry laboratories, hot cells and gloveboxes provide scientists with a controlled and safe environment to perform experiments. Working on experiments in these isolated containment cells requires scientists to be physically present. For hot cell work today, scientists manipulate equipment and radioactive material inside through a bilateral mechanical control mechanism. Motions produced outside the cell with the master control levers are mechanically transferred to the internal grippers inside the shielded containment cell. There is a growing need to have the capability to conduct experiments within these cells remotely. A simple method to enable remote manipulations within hot cell and glovebox cells is to mount two robotic arms inside a box to mimic the motions of human hands. An AR application was built in this work to allow a user wearing a Microsoft HoloLens 2 headset to teleoperate dual arm manipulators by grasping robotic end-effector digital replicas in AR from a remote location. In addition to the real-time replica of the physical robotic arms in AR, the application enables users to view a live video stream attached to the robotic arms and parse a 3D point cloud of 3D objects in their remote AR environment for better situational awareness. This work also provides users with virtual fixture to assist in manipulation and other teleoperation tasks.
521
Communicating Robot Arm Motion Intent Through Mixed Reality Head-mounted Displays
Efficient motion intent communication is necessary for safe and collaborative work environments with collocated humans and robots. Humans efficiently communicate their motion intent to other humans through gestures, gaze, and social cues. However, robots often have difficulty efficiently communicating their motion intent to humans via these methods. Many existing methods for robot motion intent communication rely on 2D displays, which require the human to continually pause their work and check a visualization. We propose a mixed reality head-mounted display visualization of the proposed robot motion over the wearer's real-world view of the robot and its environment. To evaluate the effectiveness of this system against a 2D display visualization and against no visualization, we asked 32 participants to labeled different robot arm motions as either colliding or non-colliding with blocks on a table. We found a 16% increase in accuracy with a 62% decrease in the time it took to complete the task compared to the next best system. This demonstrates that a mixed-reality HMD allows a human to more quickly and accurately tell where the robot is going to move than the compared baselines.
Liked
jechoi@andrew.cmu.edu
Communicating Robot Arm Motion Intent Through Mixed Reality Head-mounted Displays : Efficient motion intent communication is necessary for safe and collaborative work environments with collocated humans and robots. Humans efficiently communicate their motion intent to other humans through gestures, gaze, and social cues. However, robots often have difficulty efficiently communicating their motion intent to humans via these methods. Many existing methods for robot motion intent communication rely on 2D displays, which require the human to continually pause their work and check a visualization. We propose a mixed reality head-mounted display visualization of the proposed robot motion over the wearer's real-world view of the robot and its environment. To evaluate the effectiveness of this system against a 2D display visualization and against no visualization, we asked 32 participants to labeled different robot arm motions as either colliding or non-colliding with blocks on a table. We found a 16% increase in accuracy with a 62% decrease in the time it took to complete the task compared to the next best system. This demonstrates that a mixed-reality HMD allows a human to more quickly and accurately tell where the robot is going to move than the compared baselines.
1
jechoi@andrew.cmu.edu [SEP] Communicating Robot Arm Motion Intent Through Mixed Reality Head-mounted Displays : Efficient motion intent communication is necessary for safe and collaborative work environments with collocated humans and robots. Humans efficiently communicate their motion intent to other humans through gestures, gaze, and social cues. However, robots often have difficulty efficiently communicating their motion intent to humans via these methods. Many existing methods for robot motion intent communication rely on 2D displays, which require the human to continually pause their work and check a visualization. We propose a mixed reality head-mounted display visualization of the proposed robot motion over the wearer's real-world view of the robot and its environment. To evaluate the effectiveness of this system against a 2D display visualization and against no visualization, we asked 32 participants to labeled different robot arm motions as either colliding or non-colliding with blocks on a table. We found a 16% increase in accuracy with a 62% decrease in the time it took to complete the task compared to the next best system. This demonstrates that a mixed-reality HMD allows a human to more quickly and accurately tell where the robot is going to move than the compared baselines.
561
The ART of Transfer Learning: An Adaptive and Robust Pipeline
Transfer learning is an essential tool for improving the performance of primary tasks by leveraging information from auxiliary data resources. In this work, we propose Adaptive Robust Transfer Learning (ART), a flexible pipeline of performing transfer learning with generic machine learning algorithms. We establish the non-asymptotic learning theory of ART, providing a provable theoretical guarantee for achieving adaptive transfer while preventing negative transfer. Additionally, we introduce an ART-integrated-aggregating machine that produces a single final model when multiple candidate algorithms are considered. We demonstrate the promising performance of ART through extensive empirical studies on regression, classification, and sparse learning. We further present a real-data analysis for a mortality study.
Liked
zrz@andrew.cmu.edu
The ART of Transfer Learning: An Adaptive and Robust Pipeline : Transfer learning is an essential tool for improving the performance of primary tasks by leveraging information from auxiliary data resources. In this work, we propose Adaptive Robust Transfer Learning (ART), a flexible pipeline of performing transfer learning with generic machine learning algorithms. We establish the non-asymptotic learning theory of ART, providing a provable theoretical guarantee for achieving adaptive transfer while preventing negative transfer. Additionally, we introduce an ART-integrated-aggregating machine that produces a single final model when multiple candidate algorithms are considered. We demonstrate the promising performance of ART through extensive empirical studies on regression, classification, and sparse learning. We further present a real-data analysis for a mortality study.
1
zrz@andrew.cmu.edu [SEP] The ART of Transfer Learning: An Adaptive and Robust Pipeline : Transfer learning is an essential tool for improving the performance of primary tasks by leveraging information from auxiliary data resources. In this work, we propose Adaptive Robust Transfer Learning (ART), a flexible pipeline of performing transfer learning with generic machine learning algorithms. We establish the non-asymptotic learning theory of ART, providing a provable theoretical guarantee for achieving adaptive transfer while preventing negative transfer. Additionally, we introduce an ART-integrated-aggregating machine that produces a single final model when multiple candidate algorithms are considered. We demonstrate the promising performance of ART through extensive empirical studies on regression, classification, and sparse learning. We further present a real-data analysis for a mortality study.
151
APEX: Ambidextrous Dual-Arm Robotic Manipulation Using Collision-Free Generative Diffusion Models
Dexterous manipulation, particularly adept coordinating and grasping, constitutes a fundamental and indispensable capability for robots, facilitating the emulation of human-like behaviors. Integrating this capability into robots empowers them to supplement and even supplant humans in undertaking increasingly intricate tasks in both daily life and industrial settings. Unfortunately, contemporary methodologies encounter serious challenges in devising manipulation trajectories owing to the intricacies of tasks, the expansive robotic manipulation space, and dynamic obstacles. We propose a novel approach, APEX, to address all these difficulties by introducing a collision-free latent diffusion model for both robotic motion planning and manipulation. Firstly, we simplify the complexity of real-life ambidextrous dual-arm robotic manipulation tasks by abstracting them as aligning two vectors. Secondly, we devise latent diffusion models to produce a variety of robotic manipulation trajectories. Furthermore, we integrate obstacle information utilizing a classifier-guidance technique, thereby guaranteeing both the feasibility and safety of the generated manipulation trajectories. Lastly, we validate our proposed algorithm through extensive experiments conducted on the hardware platform of ambidextrous dual-arm robots. Our algorithm consistently generates successful and seamless trajectories across diverse tasks, surpassing conventional robotic motion planning algorithms. These results carry significant implications for the future design of diffusion robots, enhancing their capability to tackle more intricate robotic manipulation tasks with increased efficiency and safety. Complete video demonstrations of our experiments can be found in https://sites.google.com/view/apex-dual-arm/home.
Liked
jechoi@andrew.cmu.edu
APEX: Ambidextrous Dual-Arm Robotic Manipulation Using Collision-Free Generative Diffusion Models : Dexterous manipulation, particularly adept coordinating and grasping, constitutes a fundamental and indispensable capability for robots, facilitating the emulation of human-like behaviors. Integrating this capability into robots empowers them to supplement and even supplant humans in undertaking increasingly intricate tasks in both daily life and industrial settings. Unfortunately, contemporary methodologies encounter serious challenges in devising manipulation trajectories owing to the intricacies of tasks, the expansive robotic manipulation space, and dynamic obstacles. We propose a novel approach, APEX, to address all these difficulties by introducing a collision-free latent diffusion model for both robotic motion planning and manipulation. Firstly, we simplify the complexity of real-life ambidextrous dual-arm robotic manipulation tasks by abstracting them as aligning two vectors. Secondly, we devise latent diffusion models to produce a variety of robotic manipulation trajectories. Furthermore, we integrate obstacle information utilizing a classifier-guidance technique, thereby guaranteeing both the feasibility and safety of the generated manipulation trajectories. Lastly, we validate our proposed algorithm through extensive experiments conducted on the hardware platform of ambidextrous dual-arm robots. Our algorithm consistently generates successful and seamless trajectories across diverse tasks, surpassing conventional robotic motion planning algorithms. These results carry significant implications for the future design of diffusion robots, enhancing their capability to tackle more intricate robotic manipulation tasks with increased efficiency and safety. Complete video demonstrations of our experiments can be found in https://sites.google.com/view/apex-dual-arm/home.
1
jechoi@andrew.cmu.edu [SEP] APEX: Ambidextrous Dual-Arm Robotic Manipulation Using Collision-Free Generative Diffusion Models : Dexterous manipulation, particularly adept coordinating and grasping, constitutes a fundamental and indispensable capability for robots, facilitating the emulation of human-like behaviors. Integrating this capability into robots empowers them to supplement and even supplant humans in undertaking increasingly intricate tasks in both daily life and industrial settings. Unfortunately, contemporary methodologies encounter serious challenges in devising manipulation trajectories owing to the intricacies of tasks, the expansive robotic manipulation space, and dynamic obstacles. We propose a novel approach, APEX, to address all these difficulties by introducing a collision-free latent diffusion model for both robotic motion planning and manipulation. Firstly, we simplify the complexity of real-life ambidextrous dual-arm robotic manipulation tasks by abstracting them as aligning two vectors. Secondly, we devise latent diffusion models to produce a variety of robotic manipulation trajectories. Furthermore, we integrate obstacle information utilizing a classifier-guidance technique, thereby guaranteeing both the feasibility and safety of the generated manipulation trajectories. Lastly, we validate our proposed algorithm through extensive experiments conducted on the hardware platform of ambidextrous dual-arm robots. Our algorithm consistently generates successful and seamless trajectories across diverse tasks, surpassing conventional robotic motion planning algorithms. These results carry significant implications for the future design of diffusion robots, enhancing their capability to tackle more intricate robotic manipulation tasks with increased efficiency and safety. Complete video demonstrations of our experiments can be found in https://sites.google.com/view/apex-dual-arm/home.
494