Datasets:

dwhouse

/

arxiv_cs_abstract_20

like 0

PiCoGen2: Piano cover generation with transfer learning approach and weakly aligned data

Piano cover generation aims to create a piano cover from a pop song. Existing approaches mainly employ supervised learning and the training demands strongly-aligned and paired song-to-piano data, which is built by remapping piano notes to song audio. This would, however, result in the loss of piano information and accordingly cause inconsistencies between the original and remapped piano versions. To overcome this limitation, we propose a transfer learning approach that pre-trains our model on piano-only data and fine-tunes it on weakly-aligned paired data constructed without note remapping. During pre-training, to guide the model to learn piano composition concepts instead of merely transcribing audio, we use an existing lead sheet transcription model as the encoder to extract high-level features from the piano recordings. The pre-trained model is then fine-tuned on the paired song-piano data to transfer the learned composition knowledge to the pop song domain. Our evaluation shows that this training strategy enables our model, named PiCoGen2, to attain high-quality results, outperforming baselines on both objective and subjective metrics across five pop genres.

Single-Pixel Fluorescent Diffraction Tomography

Optical diffraction tomography is an indispensable tool for studying objects in three-dimensions due to its ability to accurately reconstruct scattering objects. Until now this technique has been limited to coherent light because spatial phase information is required to solve the inverse scattering problem. We introduce a method that extends optical diffraction tomography to imaging spatially incoherent contrast mechanisms such as fluorescent emission. Our strategy mimics the coherent scattering process with two spatially coherent illumination beams. The interferometric illumination pattern encodes spatial phase in temporal variations of the fluorescent emission, thereby allowing incoherent fluorescent emission to mimic the behavior of coherent illumination. The temporal variations permit recovery of the propagation phase, and thus the spatial distribution of incoherent fluorescent emission can be recovered with an inverse scattering model.

Curvature-Aware Derivative-Free Optimization

The paper discusses derivative-free optimization (DFO), which involves minimizing a function without access to gradients or directional derivatives, only function evaluations. Classical DFO methods, which mimic gradient-based methods, such as Nelder-Mead and direct search have limited scalability for high-dimensional problems. Zeroth-order methods have been gaining popularity due to the demands of large-scale machine learning applications, and the paper focuses on the selection of the step size $\alpha_k$ in these methods. The proposed approach, called Curvature-Aware Random Search (CARS), uses first- and second-order finite difference approximations to compute a candidate $\alpha_{+}$. We prove that for strongly convex objective functions, CARS converges linearly provided that the search direction is drawn from a distribution satisfying very mild conditions. We also present a Cubic Regularized variant of CARS, named CARS-CR, which converges in a rate of $\mathcal{O}(k^{-1})$ without the assumption of strong convexity. Numerical experiments show that CARS and CARS-CR match or exceed the state-of-the-arts on benchmark problem sets.

Ad hoc Cloud Computing: From Concept to Realization

This paper presents the first complete, integrated and end-to-end solution for ad hoc cloud computing environments. Ad hoc clouds harvest resources from existing sporadically available, non-exclusive (i.e. primarily used for some other purpose) and unreliable infrastructures. In this paper we discuss the problems ad hoc cloud computing solves and outline our architecture which is based on BOINC.

Cluster-based Contrastive Disentangling for Generalized Zero-Shot Learning

Generalized Zero-Shot Learning (GZSL) aims to recognize both seen and unseen classes by training only the seen classes, in which the instances of unseen classes tend to be biased towards the seen class. In this paper, we propose a Cluster-based Contrastive Disentangling (CCD) method to improve GZSL by alleviating the semantic gap and domain shift problems. Specifically, we first cluster the batch data to form several sets containing similar classes. Then, we disentangle the visual features into semantic-unspecific and semantic-matched variables, and further disentangle the semantic-matched variables into class-shared and class-unique variables according to the clustering results. The disentangled learning module with random swapping and semantic-visual alignment bridges the semantic gap. Moreover, we introduce contrastive learning on semantic-matched and class-unique variables to learn high intra-set and intra-class similarity, as well as inter-set and inter-class discriminability. Then, the generated visual features conform to the underlying characteristics of general images and have strong discriminative information, which alleviates the domain shift problem well. We evaluate our proposed method on four datasets and achieve state-of-the-art results in both conventional and generalized settings.

Model Predictive Control and Reinforcement Learning: A Unified Framework Based on Dynamic Programming

In this paper we describe a new conceptual framework that connects approximate Dynamic Programming (DP), Model Predictive Control (MPC), and Reinforcement Learning (RL). This framework centers around two algorithms, which are designed largely independently of each other and operate in synergy through the powerful mechanism of Newton's method. We call them the off-line training and the on-line play algorithms. The names are borrowed from some of the major successes of RL involving games; primary examples are the recent (2017) AlphaZero program (which plays chess, [SHS17], [SSS17]), and the similarly structured and earlier (1990s) TD-Gammon program (which plays backgammon, [Tes94], [Tes95], [TeG96]). In these game contexts, the off-line training algorithm is the method used to teach the program how to evaluate positions and to generate good moves at any given position, while the on-line play algorithm is the method used to play in real time against human or computer opponents. Significantly, the synergy between off-line training and on-line play also underlies MPC (as well as other major classes of sequential decision problems), and indeed the MPC design architecture is very similar to the one of AlphaZero and TD-Gammon. This conceptual insight provides a vehicle for bridging the cultural gap between RL and MPC, and sheds new light on some fundamental issues in MPC. These include the enhancement of stability properties through rollout, the treatment of uncertainty through the use of certainty equivalence, the resilience of MPC in adaptive control settings that involve changing system parameters, and the insights provided by the superlinear performance bounds implied by Newton's method.

First-Order vs. Second-Order Encodings for LTLf-to-Automata Translation

Translating formulas of Linear Temporal Logic (LTL) over finite traces, or LTLf, to symbolic Deterministic Finite Automata (DFA) plays an important role not only in LTLf synthesis, but also in synthesis for Safety LTL formulas. The translation is enabled by using MONA, a powerful tool for symbolic, BDD-based, DFA construction from logic specifications. Recent works used a first-order encoding of LTLf formulas to translate LTLf to First Order Logic (FOL), which is then fed to MONA to get the symbolic DFA. This encoding was shown to perform well, but other encodings have not been studied. Specifically, the natural question of whether second-order encoding, which has significantly simpler quantificational structure, can outperform first-order encoding remained open. In this paper we address this challenge and study second-order encodings for LTLf formulas. We first introduce a specific MSO encoding that captures the semantics of LTLf in a natural way and prove its correctness. We then explore is a Compact MSO encoding, which benefits from automata-theoretic minimization, thus suggesting a possible practical advantage. To that end, we propose a formalization of symbolic DFA in second-order logic, thus developing a novel connection between BDDs and MSO. We then show by empirical evaluations that the first-order encoding does perform better than both second-order encodings. The conclusion is that first-order encoding is a better choice than second-order encoding in LTLf-to-Automata translation.

Deep Learning Approach in Automatic Iceberg - Ship Detection with SAR Remote Sensing Data

Deep Learning is gaining traction with geophysics community to understand subsurface structures, such as fault detection or salt body in seismic data. This study describes using deep learning method for iceberg or ship recognition with synthetic aperture radar (SAR) data. Drifting icebergs pose a potential threat to activities offshore around the Arctic, including for both ship navigation and oil rigs. Advancement of satellite imagery using weather-independent cross-polarized radar has enabled us to monitor and delineate icebergs and ships, however a human component is needed to classify the images. Here we present Transfer Learning, a convolutional neural network (CNN) designed to work with a limited training data and features, while demonstrating its effectiveness in this problem. Key aspect of the approach is data augmentation and stacking of multiple outputs, resulted in a significant boost in accuracy (logarithmic score of 0.1463). This algorithm has been tested through participation at the Statoil/C-Core Kaggle competition.

Region-Aware Face Swapping

This paper presents a novel Region-Aware Face Swapping (RAFSwap) network to achieve identity-consistent harmonious high-resolution face generation in a local-global manner: \textbf{1)} Local Facial Region-Aware (FRA) branch augments local identity-relevant features by introducing the Transformer to effectively model misaligned cross-scale semantic interaction. \textbf{2)} Global Source Feature-Adaptive (SFA) branch further complements global identity-relevant cues for generating identity-consistent swapped faces. Besides, we propose a \textit{Face Mask Predictor} (FMP) module incorporated with StyleGAN2 to predict identity-relevant soft facial masks in an unsupervised manner that is more practical for generating harmonious high-resolution faces. Abundant experiments qualitatively and quantitatively demonstrate the superiority of our method for generating more identity-consistent high-resolution swapped faces over SOTA methods, \eg, obtaining 96.70 ID retrieval that outperforms SOTA MegaFS by 5.87$\uparrow$.

How wireless queues benefit from motion: an analysis of the continuum between zero and infinite mobility

This paper considers the time evolution of a queue that is embedded in a Poisson point process of moving wireless interferers. The queue is driven by an external arrival process and is subject to a time-varying service process that is a function of the SINR that it sees. Static configurations of interferers result in an infinite queue workload with positive probability. In contrast, a generic stability condition is established for the queue in the case where interferers possess any non-zero mobility that results in displacements that are both independent across interferers and oblivious to interferer positions. The proof leverages the mixing property of the Poisson point process. The effect of an increase in mobility on queueing metrics is also studied. Convex ordering tools are used to establish that faster moving interferers result in a queue workload that is smaller for the increasing-convex stochastic order. As a corollary, mean workload and mean delay decrease as network mobility increases. This stochastic ordering as a function of mobility is explained by establishing positive correlations between SINR level-crossing events at different time points, and by determining the autocorrelation function for interference and observing that it decreases with increasing mobility. System behaviour is empirically analyzed using discrete-event simulation and the performance of various mobility models is evaluated using heavy-traffic approximations.

Minimum Viable Model Estimates for Machine Learning Projects

Prioritization of machine learning projects requires estimates of both the potential ROI of the business case and the technical difficulty of building a model with the required characteristics. In this work we present a technique for estimating the minimum required performance characteristics of a predictive model given a set of information about how it will be used. This technique will result in robust, objective comparisons between potential projects. The resulting estimates will allow data scientists and managers to evaluate whether a proposed machine learning project is likely to succeed before any modelling needs to be done. The technique has been implemented into the open source application MinViME (Minimum Viable Model Estimator) which can be installed via the PyPI python package management system, or downloaded directly from the GitHub repository. Available at https://github.com/john-hawkins/MinViME

Meta Architecture Search

Neural Architecture Search (NAS) has been quite successful in constructing state-of-the-art models on a variety of tasks. Unfortunately, the computational cost can make it difficult to scale. In this paper, we make the first attempt to study Meta Architecture Search which aims at learning a task-agnostic representation that can be used to speed up the process of architecture search on a large number of tasks. We propose the Bayesian Meta Architecture SEarch (BASE) framework which takes advantage of a Bayesian formulation of the architecture search problem to learn over an entire set of tasks simultaneously. We show that on Imagenet classification, we can find a model that achieves 25.7% top-1 error and 8.1% top-5 error by adapting the architecture in less than an hour from an 8 GPU days pretrained meta-network. By learning a good prior for NAS, our method dramatically decreases the required computation cost while achieving comparable performance to current state-of-the-art methods - even finding competitive models for unseen datasets with very quick adaptation. We believe our framework will open up new possibilities for efficient and massively scalable architecture search research across multiple tasks.

Denoising LM: Pushing the Limits of Error Correction Models for Speech Recognition

Language models (LMs) have long been used to improve results of automatic speech recognition (ASR) systems, but they are unaware of the errors that ASR systems make. Error correction models are designed to fix ASR errors, however, they showed little improvement over traditional LMs mainly due to the lack of supervised training data. In this paper, we present Denoising LM (DLM), which is a $\textit{scaled}$ error correction model trained with vast amounts of synthetic data, significantly exceeding prior attempts meanwhile achieving new state-of-the-art ASR performance. We use text-to-speech (TTS) systems to synthesize audio, which is fed into an ASR system to produce noisy hypotheses, which are then paired with the original texts to train the DLM. DLM has several $\textit{key ingredients}$: (i) up-scaled model and data; (ii) usage of multi-speaker TTS systems; (iii) combination of multiple noise augmentation strategies; and (iv) new decoding techniques. With a Transformer-CTC ASR, DLM achieves 1.5% word error rate (WER) on $\textit{test-clean}$ and 3.3% WER on $\textit{test-other}$ on Librispeech, which to our knowledge are the best reported numbers in the setting where no external audio data are used and even match self-supervised methods which use external audio data. Furthermore, a single DLM is applicable to different ASRs, and greatly surpassing the performance of conventional LM based beam-search rescoring. These results indicate that properly investigated error correction models have the potential to replace conventional LMs, holding the key to a new level of accuracy in ASR systems.

Toroidal AutoEncoder

Enforcing distributions of latent variables in neural networks is an active subject. It is vital in all kinds of generative models, where we want to be able to interpolate between points in the latent space, or sample from it. Modern generative AutoEncoders (AE) like WAE, SWAE, CWAE add a regularizer to the standard (deterministic) AE, which allows to enforce Gaussian distribution in the latent space. Enforcing different distributions, especially topologically nontrivial, might bring some new interesting possibilities, but this subject seems unexplored so far. This article proposes a new approach to enforce uniform distribution on d-dimensional torus. We introduce a circular spring loss, which enforces minibatch points to be equally spaced and satisfy cyclic boundary conditions. As example of application we propose multiple-path morphing. Minimal distance geodesic between two points in uniform distribution on latent space of angles becomes a line, however, torus topology allows us to choose such lines in alternative ways, going through different edges of $[-\pi,\pi]^d$. Further applications to explore can be for example trying to learn real-life topologically nontrivial spaces of features, like rotations to automatically recognize 2D rotation of an object in picture by training on relative angles, or even 3D rotations by additionally using spherical features - this way morphing should be close to object rotation.

Basic tasks of sentiment analysis

Subjectivity detection is the task of identifying objective and subjective sentences. Objective sentences are those which do not exhibit any sentiment. So, it is desired for a sentiment analysis engine to find and separate the objective sentences for further analysis, e.g., polarity detection. In subjective sentences, opinions can often be expressed on one or multiple topics. Aspect extraction is a subtask of sentiment analysis that consists in identifying opinion targets in opinionated text, i.e., in detecting the specific aspects of a product or service the opinion holder is either praising or complaining about.

A Novel Transmission Scheme for the $K$-user Broadcast Channel with Delayed CSIT

The state-dependent $K$-user memoryless Broadcast Channel~(BC) with state feedback is investigated. We propose a novel transmission scheme and derive its corresponding achievable rate region, which, compared to some general schemes that deal with feedback, has the advantage of being relatively simple and thus is easy to evaluate. In particular, it is shown that the capacity region of the symmetric erasure BC with an arbitrary input alphabet size is achievable with the proposed scheme. For the fading Gaussian BC, we derive a symmetric achievable rate as a function of the signal-to-noise ratio~(SNR) and a small set of parameters. Besides achieving the optimal degrees of freedom at high SNR, the proposed scheme is shown, through numerical results, to outperform existing schemes from the literature in the finite SNR regime.

Squeezing, trisqueezing, and quadsqueezing in a spin-oscillator system

Quantum harmonic oscillators model a wide variety of phenomena ranging from electromagnetic fields to vibrations of atoms in molecules. Their excitations can be represented by bosons such as photons, single particles of light, or phonons, the quanta of vibrational energy. Linear interactions that only create and annihilate single bosons can generate coherent states of light or motion. Introducing nth-order nonlinear interactions, that instead involve n bosons, leads to increasingly complex quantum behaviour. For example, second-order interactions enable squeezing, used to enhance the precision of measurements beyond classical limits, while higher-order interactions create non-Gaussian states essential for continuous-variable quantum computation. However, generating nonlinear interactions is challenging, typically requiring higher-order derivatives of the driving field or specialized hardware. Hybrid systems, where linear interactions couple an oscillator to an additional spin, offer a solution and are readily available across many platforms. Here, using the spin of a single trapped ion coupled to its motion, we employ two linear interactions to demonstrate up to fourth-order bosonic interactions; we focus on generalised squeezing interactions and demonstrate squeezing, trisqueezing, and quadsqueezing. We characterise these interactions, including their spin dependence, and reconstruct the Wigner function of the resulting states. We also discuss the scaling of the interaction strength, where we drive the quadsqueezing interaction more than 100 times faster than using conventional techniques. Our method presents no fundamental limit in the interaction order n and applies to any platform supporting spin-dependent linear interactions. Strong higher-order nonlinear interactions unlock the study of fundamental quantum optics, quantum simulation, and computation in a hitherto unexplored regime.

Neuroprosthetic decoder training as imitation learning

Neuroprosthetic brain-computer interfaces function via an algorithm which decodes neural activity of the user into movements of an end effector, such as a cursor or robotic arm. In practice, the decoder is often learned by updating its parameters while the user performs a task. When the user's intention is not directly observable, recent methods have demonstrated value in training the decoder against a surrogate for the user's intended movement. We describe how training a decoder in this way is a novel variant of an imitation learning problem, where an oracle or expert is employed for supervised training in lieu of direct observations, which are not available. Specifically, we describe how a generic imitation learning meta-algorithm, dataset aggregation (DAgger, [1]), can be adapted to train a generic brain-computer interface. By deriving existing learning algorithms for brain-computer interfaces in this framework, we provide a novel analysis of regret (an important metric of learning efficacy) for brain-computer interfaces. This analysis allows us to characterize the space of algorithmic variants and bounds on their regret rates. Existing approaches for decoder learning have been performed in the cursor control setting, but the available design principles for these decoders are such that it has been impossible to scale them to naturalistic settings. Leveraging our findings, we then offer an algorithm that combines imitation learning with optimal control, which should allow for training of arbitrary effectors for which optimal control can generate goal-oriented control. We demonstrate this novel and general BCI algorithm with simulated neuroprosthetic control of a 26 degree-of-freedom model of an arm, a sophisticated and realistic end effector.

A 20 Gbps PAM4 Data Transmitter ASIC for Particle Physics Experiments

We present the design and test results of a novel data transmitter ASIC operating up to 20.48 Gbps with 4-level Pulse-Amplitude-Modulation (PAM4) for particle physics experiments. This ASIC, named GBS20, is fabricated in a 65 nm CMOS technology. Two serializers share a 5.12 GHz Phase Locked Loop (PLL) clock. The outputs from the serializers are combined into a PAM4 signal that directly drives a Vertical-Cavity-Surface-Emitting-Laser (VCSEL). The input data channels, each at 1.28 Gbps, are scrambled with an internal 27-1 Pseudo-Random Binary Sequence (PRBS), which also serves as a frame aligner. GBS20 is tested to work at 10.24 and 20.48 Gbps with a VCSEL-based Transmitter-Optical-Subassembly (TOSA). The power consumption of GBS20 is below 238 mW and reduced to 164 mW in the low-power mode.

A Conformer-based Waveform-domain Neural Acoustic Echo Canceller Optimized for ASR Accuracy

Acoustic Echo Cancellation (AEC) is essential for accurate recognition of queries spoken to a smart speaker that is playing out audio. Previous work has shown that a neural AEC model operating on log-mel spectral features (denoted "logmel" hereafter) can greatly improve Automatic Speech Recognition (ASR) accuracy when optimized with an auxiliary loss utilizing a pre-trained ASR model encoder. In this paper, we develop a conformer-based waveform-domain neural AEC model inspired by the "TasNet" architecture. The model is trained by jointly optimizing Negative Scale-Invariant SNR (SISNR) and ASR losses on a large speech dataset. On a realistic rerecorded test set, we find that cascading a linear adaptive AEC and a waveform-domain neural AEC is very effective, giving 56-59% word error rate (WER) reduction over the linear AEC alone. On this test set, the 1.6M parameter waveform-domain neural AEC also improves over a larger 6.5M parameter logmel-domain neural AEC model by 20-29% in easy to moderate conditions. By operating on smaller frames, the waveform neural model is able to perform better at smaller sizes and is better suited for applications where memory is limited.

On the Tradeoff Region of Secure Exact-Repair Regenerating Codes

We consider the $(n,k,d,\ell)$ secure exact-repair regenerating code problem, which generalizes the $(n,k,d)$ exact-repair regenerating code problem with the additional constraint that the stored file needs to be kept information-theoretically secure against an eavesdropper, who can access the data transmitted to regenerate a total of $\ell$ different failed nodes. For all known results on this problem, the achievable tradeoff regions between the normalized storage capacity and repair bandwidth have a single corner point, achieved by a scheme proposed by Shah, Rashmi and Kumar (the SRK point). Since the achievable tradeoff regions of the exact-repair regenerating code problem without any secrecy constraints are known to have multiple corner points in general, these existing results suggest a phase-change-like behavior, i.e., enforcing a secrecy constraint ($\ell\geq 1$) immediately reduces the tradeoff region to one with a single corner point. In this work, we first show that when the secrecy parameter $\ell$ is sufficiently large, the SRK point is indeed the only corner point of the tradeoff region. However, when $\ell$ is small, we show that the tradeoff region can in fact have multiple corner points. In particular, we establish a precise characterization of the tradeoff region for the $(7,6,6,1)$ problem, which has exactly two corner points. Thus, a smooth transition, instead of a phase-change-type of transition, should be expected as the secrecy constraint is gradually strengthened.

Approximation and FPT Algorithms for Finding DM-Irreducible Spanning Subgraphs

Finding a minimum-weight strongly connected spanning subgraph of an edge-weighted directed graph is equivalent to the weighted version of the well-known strong connectivity augmentation problem. This problem is NP-hard, and a simple $2$-approximation algorithm was proposed by Frederickson and J\'aj\'a (1981); surprisingly, it still achieves the best known approximation ratio in general. Also, Bang-Jensen and Yeo (2008) showed that the unweighted problem is FPT (fixed-parameter tractable) parameterized by the difference from a trivial upper bound of the optimal value. In this paper, we consider a generalization related to the Dulmage--Mendelsohn decompositions of bipartite graphs instead of the strong connectivity of directed graphs, and extend these approximation and FPT results to the generalized setting.

SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python

SciPy is an open source scientific computing library for the Python programming language. SciPy 1.0 was released in late 2017, about 16 years after the original version 0.1 release. SciPy has become a de facto standard for leveraging scientific algorithms in the Python programming language, with more than 600 unique code contributors, thousands of dependent packages, over 100,000 dependent repositories, and millions of downloads per year. This includes usage of SciPy in almost half of all machine learning projects on GitHub, and usage by high profile projects including LIGO gravitational wave analysis and creation of the first-ever image of a black hole (M87). The library includes functionality spanning clustering, Fourier transforms, integration, interpolation, file I/O, linear algebra, image processing, orthogonal distance regression, minimization algorithms, signal processing, sparse matrix handling, computational geometry, and statistics. In this work, we provide an overview of the capabilities and development practices of the SciPy library and highlight some recent technical developments.

Model-corrected learned primal-dual models for fast limited-view photoacoustic tomography

Learned iterative reconstructions hold great promise to accelerate tomographic imaging with empirical robustness to model perturbations. Nevertheless, an adoption for photoacoustic tomography is hindered by the need to repeatedly evaluate the computational expensive forward model. Computational feasibility can be obtained by the use of fast approximate models, but a need to compensate model errors arises. In this work we advance the methodological and theoretical basis for model corrections in learned image reconstructions by embedding the model correction in a learned primal-dual framework. Here, the model correction is jointly learned in data space coupled with a learned updating operator in image space within an unrolled end-to-end learned iterative reconstruction approach. The proposed formulation allows an extension to a primal-dual deep equilibrium model providing fixed-point convergence as well as reduced memory requirements for training. We provide theoretical and empirical insights into the proposed models with numerical validation in a realistic 2D limited-view setting. The model-corrected learned primal-dual methods show excellent reconstruction quality with fast inference times and thus providing a methodological basis for real-time capable and scalable iterative reconstructions in photoacoustic tomography.

CAAP: Class-Dependent Automatic Data Augmentation Based On Adaptive Policies For Time Series

Data Augmentation is a common technique used to enhance the performance of deep learning models by expanding the training dataset. Automatic Data Augmentation (ADA) methods are getting popular because of their capacity to generate policies for various datasets. However, existing ADA methods primarily focused on overall performance improvement, neglecting the problem of class-dependent bias that leads to performance reduction in specific classes. This bias poses significant challenges when deploying models in real-world applications. Furthermore, ADA for time series remains an underexplored domain, highlighting the need for advancements in this field. In particular, applying ADA techniques to vital signals like an electrocardiogram (ECG) is a compelling example due to its potential in medical domains such as heart disease diagnostics. We propose a novel deep learning-based approach called Class-dependent Automatic Adaptive Policies (CAAP) framework to overcome the notable class-dependent bias problem while maintaining the overall improvement in time-series data augmentation. Specifically, we utilize the policy network to generate effective sample-wise policies with balanced difficulty through class and feature information extraction. Second, we design the augmentation probability regulation method to minimize class-dependent bias. Third, we introduce the information region concepts into the ADA framework to preserve essential regions in the sample. Through a series of experiments on real-world ECG datasets, we demonstrate that CAAP outperforms representative methods in achieving lower class-dependent bias combined with superior overall performance. These results highlight the reliability of CAAP as a promising ADA method for time series modeling that fits for the demands of real-world applications.

Impact of spatial auditory navigation on user experience during augmented outdoor navigation tasks

The auditory sense of humans is important when it comes to navigation. The importance is especially high in cases when an object of interest is visually partly or fully covered. Interactions with users of technology are mainly focused on the visual domain of navigation tasks. This paper presents the results of a literature review and user study exploring the impact of spatial auditory navigation on user experience during an augmented outdoor navigation task. For the user test, participants used an augmented reality app guiding them to different locations with different digital augmentation. We conclude that the utilization of the auditory sense is yet still underrepresented in augmented reality applications. In the future, more usage scenarios for audio-augmented reality such as navigation will enhance user experience and interaction quality.

Coding for Additive White Noise Channels with Feedback Corrupted by Uniform Quantization or Bounded Noise

We present simple coding strategies, which are variants of the Schalkwijk-Kailath scheme, for communicating reliably over additive white noise channels in the presence of corrupted feedback. More specifically, we consider a framework comprising an additive white forward channel and a backward link which is used for feedback. We consider two types of corruption mechanisms in the backward link. The first is quantization noise, i.e., the encoder receives the quantized values of the past outputs of the forward channel. The quantization is uniform, memoryless and time invariant (that is, symbol-by-symbol scalar quantization), with bounded quantization error. The second corruption mechanism is an arbitrarily distributed additive bounded noise in the backward link. Here we allow symbol-by-symbol encoding at the input to the backward channel. We propose simple explicit schemes that guarantee positive information rate, in bits per channel use, with positive error exponent. If the forward channel is additive white Gaussian then our schemes achieve capacity, in the limit of diminishing amplitude of the noise components at the backward link, while guaranteeing that the probability of error converges to zero as a doubly exponential function of the block length. Furthermore, if the forward channel is additive white Gaussian and the backward link consists of an additive bounded noise channel, with signal-to-noise ratio (SNR) constrained symbol-by-symbol encoding, then our schemes are also capacity-achieving in the limit of high SNR.

Calibration of the GERDA experiment

The GERmanium Detector Array (GERDA) collaboration searched for neutrinoless double-$\beta$ decay in $^{76}$Ge with an array of about 40 high-purity isotopically-enriched germanium detectors. The experimental signature of the decay is a monoenergetic signal at Q$_{\beta\beta}$ = 2039.061(7)keV in the measured summed energy spectrum of the two emitted electrons. Both the energy reconstruction and resolution of the germanium detectors are crucial to separate a potential signal from various backgrounds, such as neutrino-accompanied double-$\beta$ decays allowed by the Standard Model. The energy resolution and stability were determined and monitored as a function of time using data from regular $^{228}$Th calibrations. In this work, we describe the calibration process and associated data analysis of the full GERDA dataset, tailored to preserve the excellent resolution of the individual germanium detectors when combining data over several years.

A comprehensive and biophysically detailed computational model of the whole human heart electromechanics

While ventricular electromechanics is extensively studied, four-chamber heart models have only been addressed recently; most of these works however neglect atrial contraction. Indeed, as atria are characterized by a complex physiology influenced by the ventricular function, developing computational models able to capture the physiological atrial function and atrioventricular interaction is very challenging. In this paper, we propose a biophysically detailed electromechanical model of the whole human heart that considers both atrial and ventricular contraction. Our model includes: i) an anatomically accurate whole-heart geometry; ii) a comprehensive myocardial fiber architecture; iii) a biophysically detailed microscale model for the active force generation; iv) a 0D closed-loop model of the circulatory system; v) the fundamental interactions among the different core models; vi) specific constitutive laws and model parameters for each cardiac region. Concerning the numerical discretization, we propose an efficient segregated-intergrid-staggered scheme and we employ recently developed stabilization techniques that are crucial to obtain a stable formulation in a four-chamber scenario. We are able to reproduce the healthy cardiac function for all the heart chambers, in terms of pressure-volume loops, time evolution of pressures, volumes and fluxes, and three-dimensional cardiac deformation, with unprecedented matching (to the best of our knowledge) with the expected physiology. We also show the importance of considering atrial contraction, fibers-stretch-rate feedback and suitable stabilization techniques, by comparing the results obtained with and without these features in the model. The proposed model represents the state-of-the-art electromechanical model of the iHEART ERC project and is a fundamental step toward the building of physics-based digital twins of the human heart.

Grape: Knowledge Graph Enhanced Passage Reader for Open-domain Question Answering

A common thread of open-domain question answering (QA) models employs a retriever-reader pipeline that first retrieves a handful of relevant passages from Wikipedia and then peruses the passages to produce an answer. However, even state-of-the-art readers fail to capture the complex relationships between entities appearing in questions and retrieved passages, leading to answers that contradict the facts. In light of this, we propose a novel knowledge Graph enhanced passage reader, namely Grape, to improve the reader performance for open-domain QA. Specifically, for each pair of question and retrieved passage, we first construct a localized bipartite graph, attributed to entity embeddings extracted from the intermediate layer of the reader model. Then, a graph neural network learns relational knowledge while fusing graph and contextual representations into the hidden states of the reader model. Experiments on three open-domain QA benchmarks show Grape can improve the state-of-the-art performance by up to 2.2 exact match score with a negligible overhead increase, with the same retriever and retrieved passages. Our code is publicly available at https://github.com/jumxglhf/GRAPE.

ParaDiS: Parallelly Distributable Slimmable Neural Networks

When several limited power devices are available, one of the most efficient ways to make profit of these resources, while reducing the processing latency and communication load, is to run in parallel several neural sub-networks and to fuse the result at the end of processing. However, such a combination of sub-networks must be trained specifically for each particular configuration of devices (characterized by number of devices and their capacities) which may vary over different model deployments and even within the same deployment. In this work we introduce parallelly distributable slimmable (ParaDiS) neural networks that are splittable in parallel among various device configurations without retraining. While inspired by slimmable networks allowing instant adaptation to resources on just one device, ParaDiS networks consist of several multi-device distributable configurations or switches that strongly share the parameters between them. We evaluate ParaDiS framework on MobileNet v1 and ResNet-50 architectures on ImageNet classification task and WDSR architecture for image super-resolution task. We show that ParaDiS switches achieve similar or better accuracy than the individual models, i.e., distributed models of the same structure trained individually. Moreover, we show that, as compared to universally slimmable networks that are not distributable, the accuracy of distributable ParaDiS switches either does not drop at all or drops by a maximum of 1 % only in the worst cases. Finally, once distributed over several devices, ParaDiS outperforms greatly slimmable models.

Chemical-protein Interaction Extraction via Gaussian Probability Distribution and External Biomedical Knowledge

Motivation: The biomedical literature contains a wealth of chemical-protein interactions (CPIs). Automatically extracting CPIs described in biomedical literature is essential for drug discovery, precision medicine, as well as basic biomedical research. Most existing methods focus only on the sentence sequence to identify these CPIs. However, the local structure of sentences and external biomedical knowledge also contain valuable information. Effective use of such information may improve the performance of CPI extraction. Results: In this paper, we propose a novel neural network-based approach to improve CPI extraction. Specifically, the approach first employs BERT to generate high-quality contextual representations of the title sequence, instance sequence, and knowledge sequence. Then, the Gaussian probability distribution is introduced to capture the local structure of the instance. Meanwhile, the attention mechanism is applied to fuse the title information and biomedical knowledge, respectively. Finally, the related representations are concatenated and fed into the softmax function to extract CPIs. We evaluate our proposed model on the CHEMPROT corpus. Our proposed model is superior in performance as compared with other state-of-the-art models. The experimental results show that the Gaussian probability distribution and external knowledge are complementary to each other. Integrating them can effectively improve the CPI extraction performance. Furthermore, the Gaussian probability distribution can effectively improve the extraction performance of sentences with overlapping relations in biomedical relation extraction tasks. Availability: Data and code are available at https://github.com/CongSun-dlut/CPI_extraction. Contact: yangzh@dlut.edu.cn, wangleibihami@gmail.com Supplementary information: Supplementary data are available at Bioinformatics online.

Spatial Tactile Brain-Computer Interface Paradigm Applying Vibration Stimuli to Large Areas of User's Back

We aim at an augmentation of communication abilities of amyotrophic lateral sclerosis (ALS) patients by creating a brain-computer interface (BCI) which can control a computer or other device by using only brain activity. As a method, we use a stimulus-driven BCI based on vibration stimuli delivered via a gaming pad to the user's back. We identify P300 responses from brain activity data in response to the vibration stimuli. The user's intentions are classified according to the P300 responses recorded in the EEG. From the results of the psychophysical and online BCI experiments, we are able to classify the P300 responses very accurately, which proves the effectiveness of the proposed method.

Continuing Progress on a Lattice QCD Software Infrastructure

We report on the progress of the software effort in the QCD Application Area of SciDAC. In particular, we discuss how the software developed under SciDAC enabled the aggressive exploitation of leadership computers, and we report on progress in the area of QCD software for multi-core architectures.

Recursive Introspection: Teaching Language Model Agents How to Self-Improve

A central piece in enabling intelligent agentic behavior in foundation models is to make them capable of introspecting upon their behavior, reasoning, and correcting their mistakes as more computation or interaction is available. Even the strongest proprietary large language models (LLMs) do not quite exhibit the ability of continually improving their responses sequentially, even in scenarios where they are explicitly told that they are making a mistake. In this paper, we develop RISE: Recursive IntroSpEction, an approach for fine-tuning LLMs to introduce this capability, despite prior work hypothesizing that this capability may not be possible to attain. Our approach prescribes an iterative fine-tuning procedure, which attempts to teach the model how to alter its response after having executed previously unsuccessful attempts to solve a hard test-time problem, with optionally additional environment feedback. RISE poses fine-tuning for a single-turn prompt as solving a multi-turn Markov decision process (MDP), where the initial state is the prompt. Inspired by principles in online imitation learning and reinforcement learning, we propose strategies for multi-turn data collection and training so as to imbue an LLM with the capability to recursively detect and correct its previous mistakes in subsequent iterations. Our experiments show that RISE enables Llama2, Llama3, and Mistral models to improve themselves with more turns on math reasoning tasks, outperforming several single-turn strategies given an equal amount of inference-time computation. We also find that RISE scales well, often attaining larger benefits with more capable models. Our analysis shows that RISE makes meaningful improvements to responses to arrive at the correct solution for challenging prompts, without disrupting one-turn abilities as a result of expressing more complex distributions.

Personalized Visited-POI Assignment to Individual Raw GPS Trajectories

Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (POI)) that are actually visited by users and assign visited-location information to the corresponding span of their (personal) GPS trajectories. We also introduce a novel algorithm to solve this assignment task. First, we exhaustively extract stay-points as candidates for significant locations using a variant of a conventional stay-point extraction method. Then we select significant locations and simultaneously assign visited-POIs to them by considering various aspects, which we formulate in integer linear programming. Experimental results conducted on an actual user dataset show that our method achieves higher accuracy in the visited-POI assignment task than the various cascaded procedures of conventional methods.

A Computational Approach to Aspectual Composition

In this paper, I argue, contrary to the prevailing opinion in the linguistics and philosophy literature, that a sortal approach to aspectual composition can indeed be explanatory. In support of this view, I develop a synthesis of competing proposals by Hinrichs, Krifka and Jackendoff which takes Jackendoff's cross-cutting sortal distinctions as its point of departure. To show that the account is well-suited for computational purposes, I also sketch an implemented calculus of eventualities which yields many of the desired inferences. Further details on both the model-theoretic semantics and the implementation can be found in (White, 1994).

A multi-term solution of the space-time Boltzmann equation for electrons in gaseous and liquid Argon

In a recent paper [1] the scattering and transport of excess electrons in liquid argon in the hydrodynamic regime was investigated, generalizing the seminal works of Lekner and Cohen [2,3] with modern scattering theory techniques and kinetic theory. In this paper, the discussion is extended to the non-hydrodynamic regime through the development of a full multi-term space-time solution of Boltzmann's equation for electron transport in gases and liquids using a novel operator-splitting method. A Green's function formalism is considered that enables flexible adaptation to various experimental systems. The spatio-temporal evolution of electrons in liquids in the hydrodynamic regime is studied for a benchmark model Percus-Yevick liquid as well as for liquid argon. The temporal evolution of Franck-Hertz oscillations are observed for liquids, with striking differences in the spatio-temporal development of the velocity distribution function components between the uncorrelated gas and true liquid approximations in argon. Transport properties calculated from the non-hydrodynamic theory in the long time limit, and under steady-state Townsend conditions, are benchmarked against hydrodynamic transport coefficients.

Learning to Represent Programs with Heterogeneous Graphs

Program source code contains complex structure information, which can be represented in structured data forms like trees or graphs. To acquire the structural information in source code, most existing researches use abstract syntax trees (AST). A group of works add additional edges to ASTs to convert source code into graphs and use graph neural networks to learn representations for program graphs. Although these works provide additional control or data flow information to ASTs for downstream tasks, they neglect an important aspect of structure information in AST itself: the different types of nodes and edges. In ASTs, different nodes contain different kinds of information like variables or control flow, and the relation between a node and all its children can also be different. To address the information of node and edge types, we bring the idea of heterogeneous graphs to learning on source code and present a new formula of building heterogeneous program graphs from ASTs with additional type information for nodes and edges. We use the ASDL grammar of programming language to define the node and edge types of program graphs. Then we use heterogeneous graph neural networks to learn on these graphs. We evaluate our approach on two tasks: code comment generation and method naming. Both tasks require reasoning on the semantics of complete code snippets. Experiment results show that our approach outperforms baseline models, including homogeneous graph-based models, showing that leveraging the type information of nodes and edges in program graphs can help in learning program semantics.

Robust Yet Efficient Conformal Prediction Sets

Conformal prediction (CP) can convert any model's output into prediction sets guaranteed to include the true label with any user-specified probability. However, same as the model itself, CP is vulnerable to adversarial test examples (evasion) and perturbed calibration data (poisoning). We derive provably robust sets by bounding the worst-case change in conformity scores. Our tighter bounds lead to more efficient sets. We cover both continuous and discrete (sparse) data and our guarantees work both for evasion and poisoning attacks (on both features and labels).

Faster Sorting Networks for $17$, $19$ and $20$ Inputs

We present new parallel sorting networks for $17$ to $20$ inputs. For $17, 19,$ and $20$ inputs these new networks are faster (i.e., they require less computation steps) than the previously known best networks. Therefore, we improve upon the known upper bounds for minimal depth sorting networks on $17, 19,$ and $20$ channels. The networks were obtained using a combination of hand-crafted first layers and a SAT encoding of sorting networks.

Flexible Non-interactive Short-term Implicit Certificate Generation for VANETs

A leading industry standard for secure and trusted communication in vehicular ad-hoc networks (VANETs) is the Security Credential Management System (SCMS). It uses anonymous certificates, functioning as pseudonyms, to preserve the privacy of vehicles. With the rapid development of advanced applications in VANETs, such as crowdsensing and federated learning, vehicles need to communicate with each other or infrastructures more frequently, leading to a higher demand for pseudonyms. However, the current approach of certificate provisioning in SCMS is not able to fully support pseudonyms, due to storage limitation, cost of connectivity establishment, and communication overhead of certificate downloading. To tackle this challenge, we propose a non-interactive approach for SCMS, allowing vehicles themselves to generate short-term key pairs and anonymous implicit certificates. Our evaluation and comparison with previous work show that our solution not only effectively reduces the communication cost, but also grants vehicles greater flexibility in certificate generation and use. On the technical side, to the best of our knowledge, this is the first work which (1) applies sanitizable signature for non-interactive anonymous certificate generation, and (2) is specifically designed for SCMS, which opens up possibilities for extensions and applications in industry.

Wavelet-based Heat Kernel Derivatives: Towards Informative Localized Shape Analysis

In this paper, we propose a new construction for the Mexican hat wavelets on shapes with applications to partial shape matching. Our approach takes its main inspiration from the well-established methodology of diffusion wavelets. This novel construction allows us to rapidly compute a multiscale family of Mexican hat wavelet functions, by approximating the derivative of the heat kernel. We demonstrate that it leads to a family of functions that inherit many attractive properties of the heat kernel (e.g., a local support, ability to recover isometries from a single point, efficient computation). Due to its natural ability to encode high-frequency details on a shape, the proposed method reconstructs and transfers $\delta$-functions more accurately than the Laplace-Beltrami eigenfunction basis and other related bases. Finally, we apply our method to the challenging problems of partial and large-scale shape matching. An extensive comparison to the state-of-the-art shows that it is comparable in performance, while both simpler and much faster than competing approaches.

Inner approximation algorithm for solving linear multiobjective optimization problems

Benson's outer approximation algorithm and its variants are the most frequently used methods for solving linear multiobjective optimization problems. These algorithms have two intertwined components: one-dimensional linear optimization one one hand, and a combinatorial part closely related to vertex numeration on the other. Their separation provides a deeper insight into Benson's algorithm, and points toward a dual approach. Two skeletal algorithms are defined which focus on the combinatorial part. Using different single-objective optimization problems - called oracle calls - yield different algorithms, such as a sequential convex hull algorithm, another version of Benson's algorithm with the theoretically best possible iteration count, the dual algorithm of Ehrgott, L\"ohne and Shao, and the new algorithm. The new algorithm has several advantages. First, the corresponding one-dimensional optimization problem uses the original constraints without adding any extra variables or constraints. Second, its iteration count meets the theoretically best possible one. As a dual algorithm, it is sequential: in each iteration it produces an extremal solution, thus can be aborted when a satisfactory solution is found. The Pareto front can be "probed" or "scanned" from several directions at any moment without adversely affecting the efficiency. Finally, it is well suited to handle highly degenerate problems where there are many linear dependencies among the constraints. On problems with ten or more objectives the implementation shows a significant increase in efficiency compared to Bensolve - due to the reduced number of iterations and the improved combinatorial handling.

In-Context Sharpness as Alerts: An Inner Representation Perspective for Hallucination Mitigation

Large language models (LLMs) frequently hallucinate and produce factual errors, yet our understanding of why they make these errors remains limited. In this study, we delve into the underlying mechanisms of LLM hallucinations from the perspective of inner representations, and discover a salient pattern associated with hallucinations: correct generations tend to have sharper context activations in the hidden states of the in-context tokens, compared to the incorrect ones. Leveraging this insight, we propose an entropy-based metric to quantify the ``sharpness'' among the in-context hidden states and incorporate it into the decoding process to formulate a constrained decoding approach. Experiments on various knowledge-seeking and hallucination benchmarks demonstrate our approach's consistent effectiveness, for example, achieving up to an 8.6 point improvement on TruthfulQA. We believe this study can improve our understanding of hallucinations and serve as a practical solution for hallucination mitigation.

Imitation Learning based Alternative Multi-Agent Proximal Policy Optimization for Well-Formed Swarm-Oriented Pursuit Avoidance

Multi-Robot System (MRS) has garnered widespread research interest and fostered tremendous interesting applications, especially in cooperative control fields. Yet little light has been shed on the compound ability of formation, monitoring and defence in decentralized large-scale MRS for pursuit avoidance, which puts stringent requirements on the capability of coordination and adaptability. In this paper, we put forward a decentralized Imitation learning based Alternative Multi-Agent Proximal Policy Optimization (IA-MAPPO) algorithm to provide a flexible and communication-economic solution to execute the pursuit avoidance task in well-formed swarm. In particular, a policy-distillation based MAPPO executor is firstly devised to capably accomplish and swiftly switch between multiple formations in a centralized manner. Furthermore, we utilize imitation learning to decentralize the formation controller, so as to reduce the communication overheads and enhance the scalability. Afterwards, alternative training is leveraged to compensate the performance loss incurred by decentralization. The simulation results validate the effectiveness of IA-MAPPO and extensive ablation experiments further show the performance comparable to a centralized solution with significant decrease in communication overheads.

Deep Complementary Joint Model for Complex Scene Registration and Few-shot Segmentation on Medical Images

Deep learning-based medical image registration and segmentation joint models utilize the complementarity (augmentation data or weakly supervised data from registration, region constraints from segmentation) to bring mutual improvement in complex scene and few-shot situation. However, further adoption of the joint models are hindered: 1) the diversity of augmentation data is reduced limiting the further enhancement of segmentation, 2) misaligned regions in weakly supervised data disturb the training process, 3) lack of label-based region constraints in few-shot situation limits the registration performance. We propose a novel Deep Complementary Joint Model (DeepRS) for complex scene registration and few-shot segmentation. We embed a perturbation factor in the registration to increase the activity of deformation thus maintaining the augmentation data diversity. We take a pixel-wise discriminator to extract alignment confidence maps which highlight aligned regions in weakly supervised data so the misaligned regions' disturbance will be suppressed via weighting. The outputs from segmentation model are utilized to implement deep-based region constraints thus relieving the label requirements and bringing fine registration. Extensive experiments on the CT dataset of MM-WHS 2017 Challenge show great advantages of our DeepRS that outperforms the existing state-of-the-art models.

Autonomous Spacecraft Navigation Based on Pulsar Timing Information

We discuss the possibility of an autonomous navigation system for spacecraft that is based on pulsar timing data. Pulsars are rapidly rotating neutron stars that are observable as variable celestial sources of electromagnetic radiation. Their periodic signals have timing stabilities comparable to atomic clocks and provide characteristic temporal signatures that can be used as natural navigation beacons, quite similar to the use of GPS satellites for navigation on Earth. By comparing pulse arrival times measured on-board the spacecraft with predicted pulse arrivals at some reference location, the spacecraft position can be determined autonomously with accuracies on the order of 5 kilometres. For a spacecraft at a distance of 10 astronomical units from Earth (e.g., Earth-Saturn), this means an improvement by a factor of 8 compared to conventional methods. Therefore this new technology is an alternative to standard navigation based on radio tracking by ground stations, without the disadvantages of uncertainty increasing with distance from Earth and the dependence on ground control.

Differential Evolution with Better and Nearest Option for Function Optimization

Differential evolution(DE) is a conventional algorithm with fast convergence speed. However, DE may be trapped in local optimal solution easily. Many researchers devote themselves to improving DE. In our previously work, whale swarm algorithm have shown its strong searching performance due to its niching based mutation strategy. Based on this fact, we propose a new DE algorithm called DE with Better and Nearest option (NbDE). In order to evaluate the performance of NbDE, NbDE is compared with several meta-heuristic algorithms on nine classical benchmark test functions with different dimensions. The results show that NbDE outperforms other algorithms in convergence speed and accuracy.

There Are No Post-Quantum Weakly Pseudo-Free Families in Any Nontrivial Variety of Expanded Groups

Let $\Omega$ be a finite set of finitary operation symbols and let $\mathfrak V$ be a nontrivial variety of $\Omega$-algebras. Assume that for some set $\Gamma\subseteq\Omega$ of group operation symbols, all $\Omega$-algebras in $\mathfrak V$ are groups under the operations associated with the symbols in $\Gamma$. In other words, $\mathfrak V$ is assumed to be a nontrivial variety of expanded groups. In particular, $\mathfrak V$ can be a nontrivial variety of groups or rings. Our main result is that there are no post-quantum weakly pseudo-free families in $\mathfrak V$, even in the worst-case setting and/or the black-box model. In this paper, we restrict ourselves to families $(H_d\mathbin|d\in D)$ of computational and black-box $\Omega$-algebras (where $D\subseteq\{0,1\}^*$) such that for every $d\in D$, each element of $H_d$ is represented by a unique bit string of length polynomial in the length of $d$. In our main result, we use straight-line programs to represent nontrivial relations between elements of $\Omega$-algebras. Note that under certain conditions, this result depends on the classification of finite simple groups. Also, we define and study some types of weak pseudo-freeness for families of computational and black-box $\Omega$-algebras.

Understanding the Impact of On-chip Communication on DNN Accelerator Performance

Deep Neural Networks have flourished at an unprecedented pace in recent years. They have achieved outstanding accuracy in fields such as computer vision, natural language processing, medicine or economics. Specifically, Convolutional Neural Networks (CNN) are particularly suited to object recognition or identification tasks. This, however, comes at a high computational cost, prompting the use of specialized GPU architectures or even ASICs to achieve high speeds and energy efficiency. ASIC accelerators streamline the execution of certain dataflows amenable to CNN computation that imply the constant movement of large amounts of data, thereby turning on-chip communication into a critical function within the accelerator. This paper studies the communication flows within CNN inference accelerators of edge devices, with the aim to justify current and future decisions in the design of the on-chip networks that interconnect their processing elements. Leveraging this analysis, we then qualitatively discuss the potential impact of introducing the novel paradigm of wireless on-chip network in this context.

Texture-Based Input Feature Selection for Action Recognition

The performance of video action recognition has been significantly boosted by using motion representations within a two-stream Convolutional Neural Network (CNN) architecture. However, there are a few challenging problems in action recognition in real scenarios, e.g., the variations in viewpoints and poses, and the changes in backgrounds. The domain discrepancy between the training data and the test data causes the performance drop. To improve the model robustness, we propose a novel method to determine the task-irrelevant content in inputs which increases the domain discrepancy. The method is based on a human parsing model (HP model) which jointly conducts dense correspondence labelling and semantic part segmentation. The predictions from the HP model also function as re-rendering the human regions in each video using the same set of textures to make humans appearances in all classes be the same. A revised dataset is generated for training and testing and makes the action recognition model exhibit invariance to the irrelevant content in the inputs. Moreover, the predictions from the HP model are used to enrich the inputs to the AR model during both training and testing. Experimental results show that our proposed model is superior to existing models for action recognition on the HMDB-51 dataset and the Penn Action dataset.

Privacy-preserving Scanpath Comparison for Pervasive Eye Tracking

As eye tracking becomes pervasive with screen-based devices and head-mounted displays, privacy concerns regarding eye-tracking data have escalated. While state-of-the-art approaches for privacy-preserving eye tracking mostly involve differential privacy and empirical data manipulations, previous research has not focused on methods for scanpaths. We introduce a novel privacy-preserving scanpath comparison protocol designed for the widely used Needleman-Wunsch algorithm, a generalized version of the edit distance algorithm. Particularly, by incorporating the Paillier homomorphic encryption scheme, our protocol ensures that no private information is revealed. Furthermore, we introduce a random processing strategy and a multi-layered masking method to obfuscate the values while preserving the original order of encrypted editing operation costs. This minimizes communication overhead, requiring a single communication round for each iteration of the Needleman-Wunsch process. We demonstrate the efficiency and applicability of our protocol on three publicly available datasets with comprehensive computational performance analyses and make our source code publicly accessible.

Verification of PCP-Related Computational Reductions in Coq

We formally verify several computational reductions concerning the Post correspondence problem (PCP) using the proof assistant Coq. Our verifications include a reduction of a string rewriting problem generalising the halting problem for Turing machines to PCP, and reductions of PCP to the intersection problem and the palindrome problem for context-free grammars. Interestingly, rigorous correctness proofs for some of the reductions are missing in the literature.

Fast clique minor generation in Chimera qubit connectivity graphs

The current generation of D-Wave quantum annealing processor is designed to minimize the energy of an Ising spin configuration whose pairwise interactions lie on the edges of a {\em Chimera} graph $\mathcal C_{M,N,L}$. In order to solve an Ising spin problem with arbitrary pairwise interaction structure, the corresponding graph must be minor-embedded into a Chimera graph. We define a combinatorial class of {\em native clique minors} in Chimera graphs with vertex images of uniform, near minimal size, and provide a polynomial-time algorithm that finds a maximum native clique minor in a given induced subgraph of a Chimera graph. These minors allow improvement over recent work and have immediate practical applications in the field of quantum annealing.

Approximating Robot Configuration Spaces with few Convex Sets using Clique Covers of Visibility Graphs

Many computations in robotics can be dramatically accelerated if the robot configuration space is described as a collection of simple sets. For example, recently developed motion planners rely on a convex decomposition of the free space to design collision-free trajectories using fast convex optimization. In this work, we present an efficient method for approximately covering complex configuration spaces with a small number of polytopes. The approach constructs a visibility graph using sampling and generates a clique cover of this graph to find clusters of samples that have mutual line of sight. These clusters are then inflated into large, full-dimensional, polytopes. We evaluate our method on a variety of robotic systems and show that it consistently covers larger portions of free configuration space, with fewer polytopes, and in a fraction of the time compared to previous methods.

An Optical Trap for Collisional Studies on Cold Fermionic Potassium

We report on trapping of fermionic 40K atoms in a red-detuned standing-wave optical trap, loaded from a magneto-optical trap. Typically, 10^6 atoms are loaded at a density of 10^12 cm^-3 and a temperature of 65 microK, and trapped for more than 1 s. The optical trap appears to be the proper environment for performing collisional measurements on the cold atomic sample. In particular we measure the elastic collisional rate by detecting the rethermalization following an intentional parametric heating of the atomic sample. We also measure the inelastic two-body collisional rates for unpolarized atoms in the ground hyperfine states, through detection of trap losses.

EMO: Emote Portrait Alive -- Generating Expressive Portrait Videos with Audio2Video Diffusion Model under Weak Conditions

In this work, we tackle the challenge of enhancing the realism and expressiveness in talking head video generation by focusing on the dynamic and nuanced relationship between audio cues and facial movements. We identify the limitations of traditional techniques that often fail to capture the full spectrum of human expressions and the uniqueness of individual facial styles. To address these issues, we propose EMO, a novel framework that utilizes a direct audio-to-video synthesis approach, bypassing the need for intermediate 3D models or facial landmarks. Our method ensures seamless frame transitions and consistent identity preservation throughout the video, resulting in highly expressive and lifelike animations. Experimental results demonsrate that EMO is able to produce not only convincing speaking videos but also singing videos in various styles, significantly outperforming existing state-of-the-art methodologies in terms of expressiveness and realism.

Discrete solution of the electrokinetic equations

We present a robust scheme for solving the electrokinetic equations. This goal is achieved by combining the lattice-Boltzmann method (LB) with a discrete solution of the convection-diffusion equation for the different charged and neutral species that compose the fluid. The method is based on identifying the elementary fluxes between nodes, which ensures the absence of spurious fluxes in equilibrium. We show how the model is suitable to study electro-osmotic flows. As an illustration, we show that, by introducing appropriate dynamic rules in the presence of solid interfaces, we can compute the sedimentation velocity (and hence the sedimentation potential) of a charged sphere. Our approach does not assume linearization of the Poisson-Boltzmann equation and allows us for a wide variation of the Peclet number.

Reassembling the English novel, 1789-1919

The absence of an exhaustive bibliography of novels published in the British Isles and Ireland during the 19th century blocks several lines of research in sociologically-inclined literary history and book history. Without a detailed account of novelistic production, it is difficult to characterize, for example, the population of individuals who pursued careers as novelists. This paper contributes to efforts to develop such an account by estimating yearly rates of new novel publication in the British Isles and Ireland between 1789 and 1919. This period witnessed, in aggregate, the publication of between 40,000 and 63,000 previously unpublished novels. The number of new novels published each year counts as essential information for researchers interested in understanding the development of the text industry between 1789 and 1919.

Optimal Few-GHW Linear Codes and Their Subcode Support Weight Distributions

Few-weight codes have been constructed and studied for many years, since their fascinating relations to finite geometries, strongly regular graphs and Boolean functions. Simplex codes are one-weight Griesmer $[\frac{q^k-1}{q-1},k ,q^{k-1}]_q$-linear codes and they meet all Griesmer bounds of the generalized Hamming weights of linear codes. All the subcodes with dimension $r$ of a $[\frac{q^k-1}{q-1},k ,q^{k-1}]_q$-simplex code have the same subcode support weight $\frac{q^{k-r}(q^r-1)}{q-1}$ for $1\leq r\leq k$. In this paper, we construct linear codes meeting the Griesmer bound of the $r$-generalized Hamming weight, such codes do not meet the Griesmer bound of the $j$-generalized Hamming weight for $1\leq j<r$. Moreover these codes have only few subcode support weights. The weight distribution and the subcode support weight distributions of these distance-optimal codes are determined. Linear codes constructed in this paper are natural generalizations of distance-optimal few-weight codes.

Non-Transferable Utility Coalitional Games via Mixed-Integer Linear Constraints

Coalitional games serve the purpose of modeling payoff distribution problems in scenarios where agents can collaborate by forming coalitions in order to obtain higher worths than by acting in isolation. In the classical Transferable Utility (TU) setting, coalition worths can be freely distributed amongst agents. However, in several application scenarios, this is not the case and the Non-Transferable Utility setting (NTU) must be considered, where additional application-oriented constraints are imposed on the possible worth distributions. In this paper, an approach to define NTU games is proposed which is based on describing allowed distributions via a set of mixed-integer linear constraints applied to an underlying TU game. It is shown that such games allow non-transferable conditions on worth distributions to be specified in a natural and succinct way. The properties and the relationships among the most prominent solution concepts for NTU games that hold when they are applied on (mixed-integer) constrained games are investigated. Finally, a thorough analysis is carried out to assess the impact of issuing constraints on the computational complexity of some of these solution concepts.

Augment on Manifold: Mixup Regularization with UMAP

Data augmentation techniques play an important role in enhancing the performance of deep learning models. Despite their proven benefits in computer vision tasks, their application in the other domains remains limited. This paper proposes a Mixup regularization scheme, referred to as UMAP Mixup, designed for ``on-manifold" automated data augmentation for deep learning predictive models. The proposed approach ensures that the Mixup operations result in synthesized samples that lie on the data manifold of the features and labels by utilizing a dimensionality reduction technique known as uniform manifold approximation and projection. Evaluations across diverse regression tasks show that UMAP Mixup is competitive with or outperforms other Mixup variants, show promise for its potential as an effective tool for enhancing the generalization performance of deep learning models.

Modeling film flows down a fibre influenced by nozzle geometry

We study the effects of nozzle geometry on the dynamics of thin fluid films flowing down a vertical cylindrical fibre. Recent experiments show that varying the nozzle diameter can lead to different flow regimes and droplet characteristics in the film. Using a weighted residual modeling approach, we develop a system of coupled equations that account for inertia, surface tension effects, gravity, and a film stabilization mechanism to describe both near-nozzle fluid structures and downstream bead dynamics. We report good agreement between the predicted droplet properties and the experimental data.

Thutmose Tagger: Single-pass neural model for Inverse Text Normalization

Inverse text normalization (ITN) is an essential post-processing step in automatic speech recognition (ASR). It converts numbers, dates, abbreviations, and other semiotic classes from the spoken form generated by ASR to their written forms. One can consider ITN as a Machine Translation task and use neural sequence-to-sequence models to solve it. Unfortunately, such neural models are prone to hallucinations that could lead to unacceptable errors. To mitigate this issue, we propose a single-pass token classifier model that regards ITN as a tagging task. The model assigns a replacement fragment to every input token or marks it for deletion or copying without changes. We present a dataset preparation method based on the granular alignment of ITN examples. The proposed model is less prone to hallucination errors. The model is trained on the Google Text Normalization dataset and achieves state-of-the-art sentence accuracy on both English and Russian test sets. One-to-one correspondence between tags and input words improves the interpretability of the model's predictions, simplifies debugging, and allows for post-processing corrections. The model is simpler than sequence-to-sequence models and easier to optimize in production settings. The model and the code to prepare the dataset is published as part of NeMo project.

Complex motion of precipitation bands

Formation and dynamics of an Al(OH)_3 precipitation ring is studied by diffusing NaOH into a gel containing AlCl_3. Limited feeding of the outer electrolyte (NaOH) is found to yield an intricate ring-dynamics which involves stopping and reversal of the direction of motion of the precipitation ring, and evolution into stationary multi-ring structures. A model of the ring-dynamics is developed by combining a phase separation scenario for the precipitation with the redissolution (complex formation) of the precipitate in the excess of the outer electrolyte.

A 4th-Order Particle-in-Cell Method with Phase-Space Remapping for the Vlasov-Poisson Equation

Numerical solutions to the Vlasov-Poisson system of equations have important applications to both plasma physics and cosmology. In this paper, we present a new Particle-in-Cell (PIC) method for solving this system that is 4th-order accurate in both space and time. Our method is a high-order extension of one presented previously [B. Wang, G. Miller, and P. Colella, SIAM J. Sci. Comput., 33 (2011), pp. 3509--3537]. It treats all of the stages of the standard PIC update - charge deposition, force interpolation, the field solve, and the particle push - with 4th-order accuracy, and includes a 6th-order accurate phase-space remapping step for controlling particle noise. We demonstrate the convergence of our method on a series of one- and two- dimensional electrostatic plasma test problems, comparing its accuracy to that of a 2nd-order method. As expected, the 4th-order method can achieve comparable accuracy to the 2nd-order method with many fewer resolution elements.

Reduced-order modeling of two-dimensional turbulent Rayleigh-B\'enard flow by hybrid quantum-classical reservoir computing

Two hybrid quantum-classical reservoir computing models are presented to reproduce low-order statistical properties of a two-dimensional turbulent Rayleigh-B\'enard convection flow at a Rayleigh number Ra=1e+5 and a Prandtl number Pr=10. These properties comprise the mean vertical profiles of the root mean square velocity and temperature and the turbulent convective heat flux. Both quantum algorithms differ by the arrangement of the circuit layers of the quantum reservoir, in particular the entanglement layers. The second of the two quantum circuit architectures, denoted as H2, enables a complete execution of the reservoir update inside the quantum circuit without the usage of external memory. Their performance is compared with that of a classical reservoir computing model. Therefore, all three models have to learn the nonlinear and chaotic dynamics of the turbulent flow at hand in a lower-dimensional latent data space which is spanned by the time-dependent expansion coefficients of the 16 most energetic Proper Orthogonal Decomposition (POD) modes. These training data are generated by a POD snapshot analysis from direct numerical simulations of the original turbulent flow. All reservoir computing models are operated in the reconstruction mode. We analyse different measures of the reconstruction error in dependence on the hyperparameters which are specific for the quantum cases or shared with the classical counterpart, such as the reservoir size and the leaking rate. We show that both quantum algorithms are able to reconstruct the essential statistical properties of the turbulent convection flow successfully with similar performance compared to the classical reservoir network. Most importantly, the quantum reservoirs are by a factor of 4 to 8 smaller in comparison to the classical case.

Security Evaluation for Block Scrambling-Based Image Encryption Including JPEG Distortion against Jigsaw Puzzle Solver Attacks

Encryption-then-Compression (EtC) systems have been considered for the user-controllable privacy protection of social media like Twitter. The aim of this paper is to evaluate the security of block scrambling-based encryption schemes, which have been proposed to construct EtC systems. Even though this scheme has enough key spaces against brute-force attacks, each block in encrypted images has almost the same correlation as that of original images. Therefore, it is required to consider the security from different viewpoints from number theory-based encryption methods with provable security such as RSA and AES. In this paper, we evaluate the security of encrypted images including JPEG distortion by using automatic jigsaw puzzle solvers.

Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling

The canonical approach to video-and-language learning (e.g., video question answering) dictates a neural model to learn from offline-extracted dense video features from vision models and text features from language models. These feature extractors are trained independently and usually on tasks different from the target domains, rendering these fixed features sub-optimal for downstream tasks. Moreover, due to the high computational overload of dense video features, it is often difficult (or infeasible) to plug feature extractors directly into existing approaches for easy finetuning. To provide a remedy to this dilemma, we propose a generic framework ClipBERT that enables affordable end-to-end learning for video-and-language tasks, by employing sparse sampling, where only a single or a few sparsely sampled short clips from a video are used at each training step. Experiments on text-to-video retrieval and video question answering on six datasets demonstrate that ClipBERT outperforms (or is on par with) existing methods that exploit full-length videos, suggesting that end-to-end learning with just a few sparsely sampled clips is often more accurate than using densely extracted offline features from full-length videos, proving the proverbial less-is-more principle. Videos in the datasets are from considerably different domains and lengths, ranging from 3-second generic domain GIF videos to 180-second YouTube human activity videos, showing the generalization ability of our approach. Comprehensive ablation studies and thorough analyses are provided to dissect what factors lead to this success. Our code is publicly available at https://github.com/jayleicn/ClipBERT

Skeptical Deep Learning with Distribution Correction

Recently deep neural networks have been successfully used for various classification tasks, especially for problems with massive perfectly labeled training data. However, it is often costly to have large-scale credible labels in real-world applications. One solution is to make supervised learning robust with imperfectly labeled input. In this paper, we develop a distribution correction approach that allows deep neural networks to avoid overfitting imperfect training data. Specifically, we treat the noisy input as samples from an incorrect distribution, which will be automatically corrected during our training process. We test our approach on several classification datasets with elaborately generated noisy labels. The results show significantly higher prediction and recovery accuracy with our approach compared to alternative methods.

PCPATCH: software for the topological construction of multigrid relaxation methods

Effective relaxation methods are necessary for good multigrid convergence. For many equations, standard Jacobi and Gau{\ss}-Seidel are inadequate, and more sophisticated space decompositions are required; examples include problems with semidefinite terms or saddle point structure. In this paper we present a unifying software abstraction, PCPATCH, for the topological construction of space decompositions for multigrid relaxation methods. Space decompositions are specified by collecting topological entities in a mesh (such as all vertices or faces) and applying a construction rule (such as taking all degrees of freedom in the cells around each entity). The software is implemented in PETSc and facilitates the elegant expression of a wide range of schemes merely by varying solver options at runtime. In turn, this allows for the very rapid development of fast solvers for difficult problems.

AI-enhanced on-the-fly simulation of nonlinear time-resolved spectra

Time-resolved spectroscopy is an important tool for unraveling the minute details of structural changes of molecules of biological and technological significance. The nonlinear femtosecond signals detected for such systems must be interpreted, but it is a challenging task for which theoretical simulations are often indispensable. Accurate simulations of transient-absorption or two-dimensional electronic spectra are, however, computationally very expensive, prohibiting the wider adoption of existing first-principles methods. Here, we report an AI-enhanced protocol to drastically reduce the computational cost of simulating nonlinear time-resolved electronic spectra which makes such simulations affordable for polyatomic molecules of increasing size. The protocol is based on doorway-window approach for the on-the-fly surface-hopping simulations. We show its applicability for the prototypical molecule of pyrazine for which it produces spectra with high precision with respect to ab initio reference while cutting the computational cost by at least 95% compared to pure first-principles simulations.

Phi-3 Safety Post-Training: Aligning Language Models with a "Break-Fix" Cycle

Recent innovations in language model training have demonstrated that it is possible to create highly performant models that are small enough to run on a smartphone. As these models are deployed in an increasing number of domains, it is critical to ensure that they are aligned with human preferences and safety considerations. In this report, we present our methodology for safety aligning the Phi-3 series of language models. We utilized a "break-fix" cycle, performing multiple rounds of dataset curation, safety post-training, benchmarking, red teaming, and vulnerability identification to cover a variety of harm areas in both single and multi-turn scenarios. Our results indicate that this approach iteratively improved the performance of the Phi-3 models across a wide range of responsible AI benchmarks. Finally, we include additional red teaming strategies and evaluations that were used to test the safety behavior of Phi-3.5-mini and Phi-3.5-MoE, which were optimized for multilingual capabilities.

Truly Generalizable Radiograph Segmentation with Conditional Domain Adaptation

Digitization techniques for biomedical images yield different visual patterns in radiological exams. These differences may hamper the use of data-driven approaches for inference over these images, such as Deep Neural Networks. Another noticeable difficulty in this field is the lack of labeled data, even though in many cases there is an abundance of unlabeled data available. Therefore an important step in improving the generalization capabilities of these methods is to perform Unsupervised and Semi-Supervised Domain Adaptation between different datasets of biomedical images. In order to tackle this problem, in this work we propose an Unsupervised and Semi-Supervised Domain Adaptation method for segmentation of biomedical images using Generative Adversarial Networks for Unsupervised Image Translation. We merge these unsupervised networks with supervised deep semantic segmentation architectures in order to create a semi-supervised method capable of learning from both unlabeled and labeled data, whenever labeling is available. We compare our method using several domains, datasets, segmentation tasks and traditional baselines, such as unsupervised distance-based methods and reusing pretrained models both with and without Fine-tuning. We perform both quantitative and qualitative analysis of the proposed method and baselines in the distinct scenarios considered in our experimental evaluation. The proposed method shows consistently better results than the baselines in scarce labeled data scenarios, achieving Jaccard values greater than 0.9 and good segmentation quality in most tasks. Unsupervised Domain Adaptation results were observed to be close to the Fully Supervised Domain Adaptation used in the traditional procedure of Fine-tuning pretrained networks.

Evidence for a photon mass

The author's work over the past years has indicated that the photon has a small mass $\sim 10^{-33}eV$. Recent observations from three different viewpoints -- the time lag in cosmic gamma rays with different frequencies, the observation of the spectra of blazars and an analysis of the CMB power supression from the WMAP data -- all vindicate this conclusion and remarkably, the same value.

Approximate Gram-Matrix Interpolation for Wideband Massive MU-MIMO Systems

Numerous linear and non-linear data-detection and precoding algorithms for wideband massive multi-user (MU) multiple-input multiple-output (MIMO) wireless systems that rely on orthogonal frequency-division multiplexing (OFDM) or single-carrier frequency-division multiple access (SC-FDMA) require the computation of the Gram matrix for each active subcarrier. Computing the Gram matrix for each active subcarrier, however, results in excessively high computational complexity. In this paper, we propose novel, approximate algorithms that significantly reduce the complexity of Gram-matrix computation by simultaneously exploiting correlation across subcarriers and channel hardening. We show analytically that a small fraction of Gram-matrix computations in combination with approximate interpolation schemes are sufficient to achieve near-optimal error-rate performance at low computational complexity in massive MU-MIMO systems. We also demonstrate that the proposed methods exhibit improved robustness against channel-estimation errors compared to exact Gram-matrix interpolation algorithms that typically require high computational complexity.

Computational Geometry Column 39

The resolution of a decades-old open problem is described: polygonal chains cannot lock in the plane.

More Than Meets The Eye: Semi-supervised Learning Under Non-IID Data

A common heuristic in semi-supervised deep learning (SSDL) is to select unlabelled data based on a notion of semantic similarity to the labelled data. For example, labelled images of numbers should be paired with unlabelled images of numbers instead of, say, unlabelled images of cars. We refer to this practice as semantic data set matching. In this work, we demonstrate the limits of semantic data set matching. We show that it can sometimes even degrade the performance for a state of the art SSDL algorithm. We present and make available a comprehensive simulation sandbox, called non-IID-SSDL, for stress testing an SSDL algorithm under different degrees of distribution mismatch between the labelled and unlabelled data sets. In addition, we demonstrate that simple density based dissimilarity measures in the feature space of a generic classifier offer a promising and more reliable quantitative matching criterion to select unlabelled data before SSDL training.

Beyond Observed Connections : Link Injection

In this paper, we proposed the \textit{link injection}, a novel method that helps any differentiable graph machine learning models to go beyond observed connections from the input data in an end-to-end learning fashion. It finds out (weak) connections in favor of the current task that is not present in the input data via a parametric link injection layer. We evaluate our method on both node classification and link prediction tasks using a series of state-of-the-art graph convolution networks. Results show that the link injection helps a variety of models to achieve better performances on both applications. Further empirical analysis shows a great potential of this method in efficiently exploiting unseen connections from the injected links.

Stability Analysis of Piecewise Affine Systems with Multi-model Model Predictive Control

Constrained model predictive control (MPC) is a widely used control strategy, which employs moving horizon-based on-line optimisation to compute the optimum path of the manipulated variables. Nonlinear MPC can utilize detailed models but it is computationally expensive; on the other hand linear MPC may not be adequate. Piecewise affine (PWA) models can describe the underlying nonlinear dynamics more accurately, therefore they can provide a viable trade-off through their use in multi-model linear MPC configurations, which avoid integer programming. However, such schemes may introduce uncertainty affecting the closed loop stability. In this work, we propose an input to output stability analysis for closed loop systems, consisting of PWA models, where an observer and multi-model linear MPC are applied together, under unstructured uncertainty. Integral quadratic constraints (IQCs) are employed to assess the robustness of MPC under uncertainty. We create a model pool, by performing linearisation on selected transient points. All the possible uncertainties and nonlinearities (including the controller) can be introduced in the framework, assuming that they admit the appropriate IQCs, whilst the dissipation inequality can provide necessary conditions incorporating IQCs. We demonstrate the existence of static multipliers, which can reduce the conservatism of the stability analysis significantly. The proposed methodology is demonstrated through two engineering case studies.

The motion of two identical masses connected by an ideal string symmetrically placed over a corner

We introduce a novel, two-mass system that slides up an inclined plane while its center of mass moves down. The system consists of two identical masses connected by an ideal string symmetrically placed over a corner-shaped support. This system is similar to a double-cone that rolls up an inclined set of V-shaped rails. We find the double-cone's motion easy to demonstrate but difficult to analyze. Our example here is more straightforward to follow, and the experimental observations are in good agreement with the theoretical predictions.

Confidence Intervals for Testing Disparate Impact in Fair Learning

We provide the asymptotic distribution of the major indexes used in the statistical literature to quantify disparate treatment in machine learning. We aim at promoting the use of confidence intervals when testing the so-called group disparate impact. We illustrate on some examples the importance of using confidence intervals and not a single value.

Introducing Randomized High Order Fuzzy Cognitive Maps as Reservoir Computing Models: A Case Study in Solar Energy and Load Forecasting

Fuzzy Cognitive Maps (FCMs) have emerged as an interpretable signed weighted digraph method consisting of nodes (concepts) and weights which represent the dependencies among the concepts. Although FCMs have attained considerable achievements in various time series prediction applications, designing an FCM model with time-efficient training method is still an open challenge. Thus, this paper introduces a novel univariate time series forecasting technique, which is composed of a group of randomized high order FCM models labeled R-HFCM. The novelty of the proposed R-HFCM model is relevant to merging the concepts of FCM and Echo State Network (ESN) as an efficient and particular family of Reservoir Computing (RC) models, where the least squares algorithm is applied to train the model. From another perspective, the structure of R-HFCM consists of the input layer, reservoir layer, and output layer in which only the output layer is trainable while the weights of each sub-reservoir components are selected randomly and keep constant during the training process. As case studies, this model considers solar energy forecasting with public data for Brazilian solar stations as well as Malaysia dataset, which includes hourly electric load and temperature data of the power supply company of the city of Johor in Malaysia. The experiment also includes the effect of the map size, activation function, the presence of bias and the size of the reservoir on the accuracy of R-HFCM method. The obtained results confirm the outperformance of the proposed R-HFCM model in comparison to the other methods. This study provides evidence that FCM can be a new way to implement a reservoir of dynamics in time series modelling.

Source localization using particle filtering on FPGA for robotic navigation with imprecise binary measurement

Particle filtering is a recursive Bayesian estimation technique that has gained popularity recently for tracking and localization applications. It uses Monte Carlo simulation and has proven to be a very reliable technique to model non-Gaussian and non-linear elements of physical systems. Particle filters outperform various other traditional filters like Kalman filters in non-Gaussian and non-linear settings due to their non-analytical and non-parametric nature. However, a significant drawback of particle filters is their computational complexity, which inhibits their use in real-time applications with conventional CPU or DSP based implementation schemes. This paper proposes a modification to the existing particle filter algorithm and presents a highspeed and dedicated hardware architecture. The architecture incorporates pipelining and parallelization in the design to reduce execution time considerably. The design is validated for a source localization problem wherein we estimate the position of a source in real-time using the particle filter algorithm implemented on hardware. The validation setup relies on an Unmanned Ground Vehicle (UGV) with a photodiode housing on top to sense and localize a light source. We have prototyped the design using Artix-7 field-programmable gate array (FPGA), and resource utilization for the proposed system is presented. Further, we show the execution time and estimation accuracy of the high-speed architecture and observe a significant reduction in computational time. Our implementation of particle filters on FPGA is scalable and modular, with a low execution time of about 5.62 us for processing 1024 particles and can be deployed for real-time applications.

Phase-separation transitions in asymmetric lipid bilayers

Morphological transitions of phase separation associated with the asymmetry of lipid composition were investigated using micrometer-sized vesicles of lipid bilayers made from a lipid mixture. The complete macro-phase-separated morphology undergoes a transition to a micro-phase-separation-like morphology via a lorate morphology as a metastable state. The transition leads to the emergence of monodisperse nanosized domains through repeated domain scission events. Moreover, we have numerically confirmed the transitions using the time-dependent Ginzburg-Landau model describing phase separation and the bending elastic membrane, which is quantitatively consistent with experimental results by fixing one free parameter. Our findings suggest that the local spontaneous curvature due to the asymmetric composition plays an essential role in the thermodynamic stabilization of micro-phase separation in lipid bilayers.

HiVLP: Hierarchical Vision-Language Pre-Training for Fast Image-Text Retrieval

In the past few years, the emergence of vision-language pre-training (VLP) has brought cross-modal retrieval to a new era. However, due to the latency and computation demand, it is commonly challenging to apply VLP in a real-time online retrieval system. To alleviate the defect, this paper proposes a \textbf{Hi}erarchical \textbf{V}ision-\textbf{}Language \textbf{P}re-Training (\textbf{HiVLP}) for fast Image-Text Retrieval (ITR). Specifically, we design a novel hierarchical retrieval objective, which uses the representation of different dimensions for coarse-to-fine ITR, i.e., using low-dimensional representation for large-scale coarse retrieval and high-dimensional representation for small-scale fine retrieval. We evaluate our proposed HiVLP on two popular image-text retrieval benchmarks, i.e., Flickr30k and COCO. Extensive experiments demonstrate that our HiVLP not only has fast inference speed but also can be easily scaled to large-scale ITR scenarios. The detailed results show that HiVLP is $1,427$$\sim$$120,649\times$ faster than the fusion-based model UNITER and 2$\sim$5 faster than the fastest embedding-based model LightingDot in different candidate scenarios. It also achieves about +4.9 AR on COCO and +3.8 AR on Flickr30K than LightingDot and achieves comparable performance with the state-of-the-art (SOTA) fusion-based model METER.

A Social Distancing-Based Facility Location Approach for Combating COVID-19

In this paper, we introduce and study the problem of facility location along with the notion of \emph{`social distancing'}. The input to the problem is the road network of a city where the nodes are the residential zones, edges are the road segments connecting the zones along with their respective distance. We also have the information about the population at each zone, different types of facilities to be opened and in which number, and their respective demands in each zone. The goal of the problem is to locate the facilities such that the people can be served and at the same time the total social distancing is maximized. We formally call this problem as the \textsc{Social Distancing-Based Facility Location Problem}. We mathematically quantify social distancing for a given allocation of facilities and proposed an optimization model. As the problem is \textsf{NP-Hard}, we propose a simulation-based and heuristic approach for solving this problem. A detailed analysis of both methods has been done. We perform an extensive set of experiments with synthetic datasets. From the results, we observe that the proposed heuristic approach leads to a better allocation compared to the simulation-based approach.

Deliverable navigation for multicriteria step and shoot IMRT treatment planning

We consider Pareto surface based multi-criteria optimization for step and shoot IMRT planning. By analyzing two navigation algorithms, we show both theoretically and in practice that the number of plans needed to form convex combinations of plans during navigation can be kept small (much less than the theoretical maximum number needed in general, which is equal to the number of objectives for on-surface Pareto navigation). Therefore a workable approach for directly deliverable navigation in this setting is to segment the underlying Pareto surface plans and then enforce the mild restriction that only a small number of these plans are active at any time during plan navigation, thus limiting the total number of segments used in the final plan.

Unraveling Privacy Threat Modeling Complexity: Conceptual Privacy Analysis Layers

Analyzing privacy threats in software products is an essential part of software development to ensure systems are privacy-respecting; yet it is still a far from trivial activity. While there have been many advancements in the past decade, they tend to focus on describing 'what' the threats are. What isn't entirely clear yet is 'how' to actually find these threats. Privacy is a complex domain. We propose to use four conceptual layers (feature, ecosystem, business context, and environment) to capture this privacy complexity. These layers can be used as a frame to structure and specify the privacy analysis support in a more tangible and actionable way, thereby improving applicability of the analysis process.

TMCI with Resonator Wakes

Transverse mode-coupling instability (TMCI) with a high-frequency resonator wake is examined by the Nested Head-Tail Vlasov solver (NHT), where a Gaussian bunch in a parabolic potential (GP model) is represented by concentric rings in the longitudinal phase space. It is shown that multiple mode couplings and decouplings make impossible an unambiguous definition of the threshold, unless Landau damping is taken into account. To address this problem, instead of a single instability threshold, an interval of thresholds is suggested, bounded by the low and high intensity ones. For the broadband impedance model, the high intensity threshold is shown to follow Zotter's scaling, but smaller by about a factor of two. The same scaling, this time smaller than Zotter's by a factor of four, is found for the ABS model (Air Bag Square well).

Multiorbital Quantum Impurity Solver for General Interactions and Hybridizations

We present a numerically exact Inchworm Monte Carlo method for equilibrium multiorbital quantum impurity problems with general interactions and hybridizations. We show that the method, originally developed to overcome the dynamical sign problem in certain real-time propagation problems, can also overcome the sign problem as a function of temperature for equilibrium quantum impurity models. This is shown in several cases where the current method of choice, the continuous-time hybridization expansion, fails due to the sign problem. Our method therefore enables simulations of impurity problems as they appear in embedding theories without further approximations, such as the truncation of the hybridization or interaction structure or a discretization of the impurity bath with a set of discrete energy levels, and eliminates a crucial bottleneck in the simulation of ab initio embedding problems.

Unfolding the procedure of characterizing recorded ultra low frequency, kHZ and MHz electromagetic anomalies prior to the L'Aquila earthquake as pre-seismic ones. Part I

Ultra low frequency, kHz and MHz electromagnetic anomalies were recorded prior to the L'Aquila catastrophic earthquake that occurred on April 6, 2009. The main aims of this contribution are: (i) To suggest a procedure for the designation of detected EM anomalies as seismogenic ones. We do not expect to be possible to provide a succinct and solid definition of a pre-seismic EM emission. Instead, we attempt, through a multidisciplinary analysis, to provide elements of a definition. (ii) To link the detected MHz and kHz EM anomalies with equivalent last stages of the L'Aquila earthquake preparation process. (iii) To put forward physically meaningful arguments to support a way of quantifying the time to global failure and the identification of distinguishing features beyond which the evolution towards global failure becomes irreversible. The whole effort is unfolded in two consecutive parts. We clarify we try to specify not only whether or not a single EM anomaly is pre-seismic in itself, but mainly whether a combination of kHz, MHz, and ULF EM anomalies can be characterized as pre-seismic one.

BLOFF: A Blockchain based Forensic Model in IoT

In this era of explosive growth in technology, the internet of things (IoT) has become the game changer when we consider technologies like smart homes and cities, smart energy, security and surveillance, and healthcare. The numerous benefits provided by IoT have become attractive technologies for users and cybercriminals. Cybercriminals of today have the tools and the technology to deploy millions of sophisticated attacks. These attacks need to be investigated; this is where digital forensics comes into play. However, it is not easy to conduct a forensic investigation in IoT systems because of the heterogeneous nature of the IoT environment. Additionally, forensic investigators mostly rely on evidence from service providers, a situation that can lead to evidence contamination. To solve this problem, the authors proposed a blockchain-based IoT forensic model that prevents the admissibility of tampered logs into evidence.

Capacity-achieving Sparse Superposition Codes via Approximate Message Passing Decoding

Sparse superposition codes were recently introduced by Barron and Joseph for reliable communication over the AWGN channel at rates approaching the channel capacity. The codebook is defined in terms of a Gaussian design matrix, and codewords are sparse linear combinations of columns of the matrix. In this paper, we propose an approximate message passing decoder for sparse superposition codes, whose decoding complexity scales linearly with the size of the design matrix. The performance of the decoder is rigorously analyzed and it is shown to asymptotically achieve the AWGN capacity with an appropriate power allocation. Simulation results are provided to demonstrate the performance of the decoder at finite blocklengths. We introduce a power allocation scheme to improve the empirical performance, and demonstrate how the decoding complexity can be significantly reduced by using Hadamard design matrices.

ContourDiff: Unpaired Image Translation with Contour-Guided Diffusion Models

Accurately translating medical images across different modalities (e.g., CT to MRI) has numerous downstream clinical and machine learning applications. While several methods have been proposed to achieve this, they often prioritize perceptual quality with respect to output domain features over preserving anatomical fidelity. However, maintaining anatomy during translation is essential for many tasks, e.g., when leveraging masks from the input domain to develop a segmentation model with images translated to the output domain. To address these challenges, we propose ContourDiff, a novel framework that leverages domain-invariant anatomical contour representations of images. These representations are simple to extract from images, yet form precise spatial constraints on their anatomical content. We introduce a diffusion model that converts contour representations of images from arbitrary input domains into images in the output domain of interest. By applying the contour as a constraint at every diffusion sampling step, we ensure the preservation of anatomical content. We evaluate our method by training a segmentation model on images translated from CT to MRI with their original CT masks and testing its performance on real MRIs. Our method outperforms other unpaired image translation methods by a significant margin, furthermore without the need to access any input domain information during training.

Mpemba Effect, Shechtman's Quasicrystals and Students' Exploring Activities

In the 1960s, Tanzanian student Erasto Mpemba and his teacher published an article with the title "Cool" in the journal Physics Education (Mpemba, E. B. - Osborne, D. G.: Cool?. In: Physics Education, vol.4, 1969, pp. 172-175.). In this article they claimed that hot water freezes faster than cold water. The article raised not only a wave of discussions, and other articles about this topic, but also a whole series of new experiments, which should verify this apparent thermodynamic absurdity and find an adequate explanation. Here we give a review with references to explanations and we bring some proposals for experimental student work in this area. We introduce Mpemba Effect not only as a paradoxical physics phenomenon, but we shall present a strong educational message that the Mpemba story brings to the teachers and their students. This message also creates a bridge between this phenomenon and the discovery for which the 2011 Nobel Prize in Chemistry was awarded. It leads to critical adoption of traditional knowledge and encourages resilience in investigative exploration of new things.

A Fast Eigen Solution for Homogeneous Quadratic Minimization with at most Three Constraints

We propose an eigenvalue based technique to solve the Homogeneous Quadratic Constrained Quadratic Programming problem (HQCQP) with at most 3 constraints which arise in many signal processing problems. Semi-Definite Relaxation (SDR) is the only known approach and is computationally intensive. We study the performance of the proposed fast eigen approach through simulations in the context of MIMO relays and show that the solution converges to the solution obtained using the SDR approach with significant reduction in complexity.

The diurnal cycle and temporal trends of surface winds

Winds play an essential role in the climate system. In this study, we analyze the global pattern of the diurnal cycle of surface (10 m) winds from the ERA5 reanalysis data. We find that over the land and especially over sand dune regions, the maximal wind speed and wind drift potential (DP) occur during the hours around midday. However, over the ocean, the wind also peaks at night. Using the sensible heat flux, we show that the weaker winds over land at night are due to a nocturnal cooling that decouples upper atmospheric levels and their associated stronger winds from the surface -- nocturnal cooling is much smaller over the ocean. We also analyze wind data from more than 400 meteorological stations in the USA and find a similar diurnal trend as in the reanalysis data. The timing (during the day) of the maximum wind speed has not varied much over the past 70 years. Yet, the wind speed, wind power, and wind drift potential exhibit significant increases with time over the ocean and, to a much lesser degree, over the land and sand dune regions. We compare the USA and Europe DP and wind speed of the ERA5 to that of meteorological stations and find that the ERA5 significantly underestimates real winds; however, the temporal patterns of the two are similar.

GSTran: Joint Geometric and Semantic Coherence for Point Cloud Segmentation

Learning meaningful local and global information remains a challenge in point cloud segmentation tasks. When utilizing local information, prior studies indiscriminately aggregates neighbor information from different classes to update query points, potentially compromising the distinctive feature of query points. In parallel, inaccurate modeling of long-distance contextual dependencies when utilizing global information can also impact model performance. To address these issues, we propose GSTran, a novel transformer network tailored for the segmentation task. The proposed network mainly consists of two principal components: a local geometric transformer and a global semantic transformer. In the local geometric transformer module, we explicitly calculate the geometric disparity within the local region. This enables amplifying the affinity with geometrically similar neighbor points while suppressing the association with other neighbors. In the global semantic transformer module, we design a multi-head voting strategy. This strategy evaluates semantic similarity across the entire spatial range, facilitating the precise capture of contextual dependencies. Experiments on ShapeNetPart and S3DIS benchmarks demonstrate the effectiveness of the proposed method, showing its superiority over other algorithms. The code is available at https://github.com/LAB123-tech/GSTran.