abstract
stringlengths
6
6.09k
id
stringlengths
9
16
time
int64
725k
738k
In the past years, the interest in the laser-driven acceleration of heavy ions in the mass range of A ~ 200 has been increasing due to promising application ideas like the fission-fusion nuclear reaction mechanism, aiming at the production of neutron-rich isotopes relevant for the astrophysical r-process nucleosynthesis. In this paper, we report on the laser acceleration of gold ions to beyond 7 MeV/u, exceeding for the first time an important prerequisite for this nuclear reaction scheme. Moreover, the gold ion charge states have been detected with an unprecedented resolution, which enables the separation of individual charge states up to 4 MeV/u. The recorded charge-state distributions show a remarkable dependency on the target foil thickness and differ from simulations, lacking a straight-forward explanation by the established ionization models.
2104.14520
737,909
We report complex magnetic, magnetoresistance (MR) and magnetocaloric properties of Gd4RhAl and Tb4RhAl forming in the Gd4RhIn type cubic structure. Though the synthesis of the compounds was reported long ago, to our knowledge, no attempt was made to investigate the properties of these compounds. The present results of ac and dc magnetization, electrical resistivity and heat-capacity measurements down to 1.8 K establish that these compounds undergo antiferromagnetic order initially, followed by complex spin-glass features with decreasing temperature. These characteristic temperatures are: For Gd case, TN is about 46K and TG is about 21 K, and for Tb, about 32 and 28 K respectively. Additionally, there are field induced magnetic effects, interestingly leading to non-monotonic variations in MR. There is a significant MR over a wide temperature range above TN, similar to the behavior of magnetocaloric effect (MCE) as measured by isothermal entropy change (DeltaS). An intriguing finding we made is that DeltaS at the onset of magnetic order is significantly larger for the Tb compound than that observed for the Gd analogue near its TN. On the basis of this observation in a cubic material, we raise a question whether aspherical nature of the 4f orbital can play a role to enhance MCE under favorable circumstances, a clue that could be useful to find materials for magnetocaloric applications.
2104.14521
737,909
We address the problem of decoding video file fragments when the necessary encoding parameters are missing. With this objective, we propose a method that automatically generates H.264 video headers containing these parameters and extracts coded pictures in the partially available compressed video data. To accomplish this, we examined a very large corpus of videos to learn patterns of encoding settings commonly used by encoders and created a parameter dictionary. Further, to facilitate a more efficient search our method identifies characteristics of a coded bitstream to discriminate the entropy coding mode. It also utilizes the application logs created by the decoder to identify correct parameter values. Evaluation of the effectiveness of the proposed method on more than 55K videos with diverse provenance shows that it can generate valid headers on average in 11.3 decoding trials per video. This result represents an improvement by more than a factor of 10 over the conventional approach of video header stitching to recover video file fragments.
2104.14522
737,909
Finding an effective formula for describing a discriminant of a quadrinomial (a formula which can be easily computed for high values of degrees of quadrinomials) is a difficult problem. In 2018 Otake and Shaska using advanced matrix operations found an explicit expression of $\Delta(x^n+t(x^2+ax+b))$. In this paper we focus on deriving similar results, taking advantage of alternative elementary approach, for quadrinomials of the form $x^n+ax^k+bx+c$, where $ k \in \{2,3,n-1\}$. Moreover, we make some notes about $\Delta(x^{2n}+ax^n+bx^l+c)$ such that $n>2l$.
2104.14523
737,909
In 1957 Feynman suggested that the quantum/classical character of gravity can be assessed by the presence/absence of entanglement between gravitationally interacting test masses. However, in all proposed experimental realisations using matter-wave interferometry the extreme weakness of this interaction requires pure initial states with extreme squeezing to achieve measurable entanglement for reasonable interaction times. In practice, the systems that can be prepared in such nonclassical states are limited to small masses, which in turn limits the rate at which they get entangled. Here we address this key challenge - the weakness of gravitational interaction - by using a massive body as an amplifying mediator of gravitational interaction between two test-systems. Our analysis shows, that this results in an effective interaction between the two test-systems that grows with the mass of the mediator and is independent of its initial state and, therefore, its temperature. This greatly reduces the requirement on the mass and degree of delocalization of the test systems and, while still highly challenging, brings experiments on gravitational source masses a step closer to reality.
2104.14524
737,909
We propose a change-point detection method for large scale multiple testing problems with data having clustered signals. Unlike the classic change-point setup, the signals can vary in size within a cluster. The clustering structure on the signals enables us to effectively delineate the boundaries between signal and non-signal segments. New test statistics are proposed for observations from one and/or multiple realizations. Their asymptotic distributions are derived. We also study the associated variance estimation problem. We allow the variances to be heteroscedastic in the multiple realization case, which substantially expands the applicability of the proposed method. Simulation studies demonstrate that the proposed approach has a favorable performance. Our procedure is applied to {an array based Comparative Genomic Hybridization (aCGH)} dataset.
2104.14525
737,909
Tensors, which provide a powerful and flexible model for representing multi-attribute data and multi-way interactions, play an indispensable role in modern data science across various fields in science and engineering. A fundamental task is to faithfully recover the tensor from highly incomplete measurements in a statistically and computationally efficient manner. Harnessing the low-rank structure of tensors in the Tucker decomposition, this paper develops a scaled gradient descent (ScaledGD) algorithm to directly recover the tensor factors with tailored spectral initializations, and shows that it provably converges at a linear rate independent of the condition number of the ground truth tensor for two canonical problems -- tensor completion and tensor regression -- as soon as the sample size is above the order of $n^{3/2}$ ignoring other dependencies, where $n$ is the dimension of the tensor. This leads to an extremely scalable approach to low-rank tensor estimation compared with prior art, which suffers from at least one of the following drawbacks: extreme sensitivity to ill-conditioning, high per-iteration costs in terms of memory and computation, or poor sample complexity guarantees. To the best of our knowledge, ScaledGD is the first algorithm that achieves near-optimal statistical and computational complexities simultaneously for low-rank tensor completion with the Tucker decomposition. Our algorithm highlights the power of appropriate preconditioning in accelerating nonconvex statistical estimation, where the iteration-varying preconditioners promote desirable invariance properties of the trajectory with respect to the underlying symmetry in low-rank tensor factorization.
2104.14526
737,909
We propose to assess the fairness of personalized recommender systems in the sense of envy-freeness: every (group of) user(s) should prefer their recommendations to the recommendations of other (groups of) users. Auditing for envy-freeness requires probing user preferences to detect potential blind spots, which may deteriorate recommendation performance. To control the cost of exploration, we propose an auditing algorithm based on pure exploration and conservative constraints in multi-armed bandits. We study, both theoretically and empirically, the trade-offs achieved by this algorithm.
2104.14527
737,909
For deep learning methods applied to the diagnosis of gastric cancer intelligently, existing methods concentrate more on Convolutional Neural Networks (CNN) but no approaches are available using Visual Transformer (VT). VT's efficient and stable deep learning models with the most recent application in the field of computer vision, which is capable of improving the recognition of global information in images. In this paper, a multi-scale visual transformer model (GasHis-Transformer) is proposed for a gastric histopathology image classification (GHIC) task, which enables the automatic classification of gastric histological images of abnormal and normal cancer by obtained by optical microscopy to facilitate the medical work of histopathologists. This GasHis-Transformer model is built on two fundamental modules, including a global information module (GIM) and a local information module (LIM). In the experiment, an open source hematoxylin and eosin (H&E) stained gastric histopathology dataset with 280 abnormal or normal images are divided into training, validation, and test sets at a ratio of 1:1:2 first. Then, GasHis-Transformer obtains precision, recall, F1-score, and accuracy on the testing set of 98.0%, 100.0%, 96.0%, and 98.0%. Furthermore, a contrast experiment also tests the generalization ability of the proposed GatHis-Transformer model with a lymphoma image dataset including 374 images and a breast cancer dataset including 1390 images in two extended experiments and achieves an accuracy of 83.9% and 89.4%, respectively. Finally, GasHis-Transformer model demonstrates high classification performance and shows its effectiveness and enormous potential in GHIC tasks.
2104.14528
737,909
We introduce a family of order $N\in \mathbb{N}$ Lax matrices that is indexed by the natural number $k\in \{1,\ldots,N-1\}.$ For each value of $k$ they serve as strong Lax matrices of a hierarchy of integrable difference systems in edge variables that in turn lead to hierarchies of integrable difference systems in vertex variables or in a combination of edge and vertex variables. Furthermore, the entries of the Lax matrices are considered as elements of a division ring, so we obtain hierarchies of discrete integrable systems extended in the non-commutative domain.
2104.14529
737,909
We introduce a two-parameter function $\phi_{q_+,q_-}$ on the infinite hyperoctahedral group, which is a bivariate refinement of the reflection length keeping track of the long and the short reflections separately. We provide a complete characterization of the parameters $q_+,q_-$ when the signed reflection function $\phi_{q_+,q_-}$ is positive definite and we prove that this condition holds if and only if $\phi_{q_+,q_-}$ is an extreme character of the infinite hyperoctahedral group. We construct the corresponding representations as a natural action of the hyperoctahedral group $B(n)$ on the tensor product of $n$ copies of a vector space, which gives a two-parameter analog of the classical construction of Schur--Weyl. We apply our characterization to construct a cyclic Fock space of type B which generalizes the one-parameter construction in type A found previously by Bo\.zejko and Guta. We also construct a new cyclic Gaussian operator of type B and we relate its moments with the Askey--Wilson--Kerov distribution by using the notion of cycles on pair-partitions, which we introduce here.
2104.14530
737,909
In this paper we address a variant of the Kazhdan-Lusztig non-degeneracy conjecture posed by Gedeon, Proudfoot and Young. We prove that if $M$ has a free basis (something that conjecturally asymptotically all matroids are expected to possess), then $M$ is non-degenerate. To this end, we study the behavior of Kazhdan-Lusztig polynomials of matroids with respect to the operation of circuit-hyperplane relaxation. This yields a family of polynomials that relate the Kazhdan-Lusztig, the inverse Kazhdan-Lusztig and the $Z$-polynomial of a matroid with those of its relaxations and do not depend on the matroid. As an application of our results, we deduce that uniform matroids maximize coefficient-wise the Kazhdan-Lusztig polynomials, inverse Kazhdan-Lusztig polynomials and the $Z$-polynomials, when restricted to sparse paving matroids.
2104.14531
737,909
The mass media play at least five basic functions which include news dissemination, surveillance of the environment, correlation of the components of the society, entertainment and transmission of social heritage. Sometimes, disruptions and impairments do occur in the performance of these roles and some of these basic functions become dysfunctions, which turn the media into purveyor of negative values. The present study investigates how popular the Nigerian TV reality show, Big Brother Naija BBN, is perceived by its viewers. Three hundred heavy viewers of the program were surveyed from Lagos and Ede, South-West Nigeria, and their opinions and attitudes were sought regarding, why they like or dislike the program; the gratifications that those who like the program derive and whether the BBN, as media content, is generally functional or dysfunctional to the society. Sixty six per cent 66 33.7 of respondents like the program because it entertains. Half of the respondents, 99 50.5 dislike immoral aspects of the program. The viewers affirm that the eviction part of the program was their highest form of gratification. Most respondents, despite public outcry against the program, consider the program to be functional. Findings reinforce the postulation that TV viewers are not passive consumers of media contents.
2104.14532
737,909
We propose an optimization scheme for ground-state cooling of a mechanical mode by coupling to a general three-level system. We formulate the optimization scheme, using the master equation approach, over a broad range of system parameters including detunings, decay rates, coupling strengths, and pumping rate. We implement the optimization scheme on three physical systems: a colloidal quantum dot coupled to its confined phonon mode, a polariton coupled to a mechanical resonator mode, and a coupled-cavity system coupled to a mechanical resonator mode. These three physical systems span a broad range of mechanical mode frequencies, coupling rates, and decay rates. Our optimization scheme lowers the stead-state phonon number in all three cases by orders of magnitude. We also calculate the net cooling rate by estimating the phonon decay rate and show that the optimized system parameters also result in efficient cooling. The proposed optimization scheme can be readily extended to any generic driven three-level system coupled to a mechanical mode.
2104.14533
737,909
Balancing and push-recovery are essential capabilities enabling humanoid robots to solve complex locomotion tasks. In this context, classical control systems tend to be based on simplified physical models and hard-coded strategies. Although successful in specific scenarios, this approach requires demanding tuning of parameters and switching logic between specifically-designed controllers for handling more general perturbations. We apply model-free Deep Reinforcement Learning for training a general and robust humanoid push-recovery policy in a simulation environment. Our method targets high-dimensional whole-body humanoid control and is validated on the iCub humanoid. Reward components incorporating expert knowledge on humanoid control enable fast learning of several robust behaviors by the same policy, spanning the entire body. We validate our method with extensive quantitative analyses in simulation, including out-of-sample tasks which demonstrate policy robustness and generalization, both key requirements towards real-world robot deployment.
2104.14534
737,909
Anomaly detection, the task of identifying unusual samples in data, often relies on a large set of training samples. In this work, we consider the setting of few-shot anomaly detection in images, where only a few images are given at training. We devise a hierarchical generative model that captures the multi-scale patch distribution of each training image. We further enhance the representation of our model by using image transformations and optimize scale-specific patch-discriminators to distinguish between real and fake patches of the image, as well as between different transformations applied to those patches. The anomaly score is obtained by aggregating the patch-based votes of the correct transformation across scales and image regions. We demonstrate the superiority of our method on both the one-shot and few-shot settings, on the datasets of Paris, CIFAR10, MNIST and FashionMNIST as well as in the setting of defect detection on MVTec. In all cases, our method outperforms the recent baseline methods.
2104.14535
737,909
#StopAsianHate and #StopAAPIHate are two of the most commonly used hashtags that represent the current movement to end hate crimes against the Asian American and Pacific Islander community. We conduct a social media study of public opinion on the #StopAsianHate and #StopAAPIHate movement based on 46,058 Twitter users across 30 states in the United States ranging from March 18 to April 11, 2021. The movement attracts more participation from women, younger adults, Asian and Black communities. 51.56% of the Twitter users show direct support, 18.38% are news about anti-Asian hate crimes, while 5.43% show a negative attitude towards the movement. Public opinion varies across user characteristics. Furthermore, among the states with most racial bias motivated hate crimes, the negative attitude towards the #StopAsianHate and #StopAAPIHate movement is the weakest. To our best knowledge, this is the first large-scale social media-based study to understand public opinion on the #StopAsianHate and #StopAAPIHate movement. We hope our study can provide insights and promote research on anti-Asian hate crimes, and ultimately help address such a serious societal issue for the common benefits of all communities.
2104.14536
737,909
While learned video codecs have demonstrated great promise, they have yet to achieve sufficient efficiency for practical deployment. In this work, we propose several novel ideas for learned video compression which allow for improved performance for the low-latency mode (I- and P-frames only) along with a considerable increase in computational efficiency. In this setting, for natural videos our approach compares favorably across the entire R-D curve under metrics PSNR, MS-SSIM and VMAF against all mainstream video standards (H.264, H.265, AV1) and all ML codecs. At the same time, our approach runs at least 5x faster and has fewer parameters than all ML codecs which report these figures. Our contributions include a flexible-rate framework allowing a single model to cover a large and dense range of bitrates, at a negligible increase in computation and parameter count; an efficient backbone optimized for ML-based codecs; and a novel in-loop flow prediction scheme which leverages prior information towards more efficient compression. We benchmark our method, which we call ELF-VC (Efficient, Learned and Flexible Video Coding) on popular video test sets UVG and MCL-JCV under metrics PSNR, MS-SSIM and VMAF. For example, on UVG under PSNR, it reduces the BD-rate by 44% against H.264, 26% against H.265, 15% against AV1, and 35% against the current best ML codec. At the same time, on an NVIDIA Titan V GPU our approach encodes/decodes VGA at 49/91 FPS, HD 720 at 19/35 FPS, and HD 1080 at 10/18 FPS.
2104.14335
737,909
Though machine learning models are achieving great success, ex-tensive studies have exposed their disadvantage of inheriting latent discrimination and societal bias from the training data, which hinders their adoption on high-state applications. Thus, many efforts have been taken for developing fair machine learning models. Most of them require that sensitive attributes are available during training to learn fair models. However, in many real-world applications, it is usually infeasible to obtain the sensitive attribute due to privacy or legal issues, which challenges existing fair classifiers. Though the sensitive attribute of each data sample is unknown, we observe that there are usually some non-sensitive features in the training data that are highly correlated with sensitive attributes, which can be used to alleviate the bias. Therefore, in this paper, we study a novel problem of exploring features that are highly correlated with sensitive attributes for learning fair and accurate classifier without sensitive attributes. We theoretically show that by minimizing the correlation between these related features and model prediction, we can learn a fair classifier. Based on this motivation, we propose a novel framework which simultaneously uses these related features for accurate prediction and regularizing the model to be fair. In addition, the model can dynamically adjust the importance weight of each related feature to balance the contribution of the feature on model classification and fairness. Experimental results on real-world datasets demonstrate the effectiveness of the proposed model for learning fair models with high classification accuracy.
2104.14537
737,909
We consider the distributed training of large-scale neural networks that serve as PDE solvers producing full field outputs. We specifically consider neural solvers for the generalized 3D Poisson equation over megavoxel domains. A scalable framework is presented that integrates two distinct advances. First, we accelerate training a large model via a method analogous to the multigrid technique used in numerical linear algebra. Here, the network is trained using a hierarchy of increasing resolution inputs in sequence, analogous to the 'V', 'W', 'F', and 'Half-V' cycles used in multigrid approaches. In conjunction with the multi-grid approach, we implement a distributed deep learning framework which significantly reduces the time to solve. We show the scalability of this approach on both GPU (Azure VMs on Cloud) and CPU clusters (PSC Bridges2). This approach is deployed to train a generalized 3D Poisson solver that scales well to predict output full-field solutions up to the resolution of 512x512x512 for a high dimensional family of inputs.
2104.14538
737,909
Model bias is an inherent limitation of the current dominant approach to optimal quantum control, which relies on a system simulation for optimization of control policies. To overcome this limitation, we propose a circuit-based approach for training a reinforcement learning agent on quantum control tasks in a model-free way. Given a continuously parameterized control circuit, the agent learns its parameters through trial-and-error interaction with the quantum system, using measurements as the only source of information about the quantum state. By focusing on the task of quantum state preparation in a harmonic oscillator coupled to an ancilla qubit, we show how to reward the learning agent using measurements of experimentally available observables. We demonstrate by numerical simulations preparation of arbitrary states using both open- and closed-loop control through adaptive quantum feedback. Our work is of immediate relevance to superconducting circuits and trapped ions platforms where such training can be implemented real-time in an experiment, allowing complete elimination of model bias and the adaptation of quantum control policies to the specific system in which they are deployed.
2104.14539
737,909
Self-supervised monocular depth estimation networks are trained to predict scene depth using nearby frames as a supervision signal during training. However, for many applications, sequence information in the form of video frames is also available at test time. The vast majority of monocular networks do not make use of this extra signal, thus ignoring valuable information that could be used to improve the predicted depth. Those that do, either use computationally expensive test-time refinement techniques or off-the-shelf recurrent networks, which only indirectly make use of the geometric information that is inherently available. We propose ManyDepth, an adaptive approach to dense depth estimation that can make use of sequence information at test time, when it is available. Taking inspiration from multi-view stereo, we propose a deep end-to-end cost volume based approach that is trained using self-supervision only. We present a novel consistency loss that encourages the network to ignore the cost volume when it is deemed unreliable, e.g. in the case of moving objects, and an augmentation scheme to cope with static cameras. Our detailed experiments on both KITTI and Cityscapes show that we outperform all published self-supervised baselines, including those that use single or multiple frames at test time.
2104.14540
737,909
An equation has been derived to predict unit cell volume of high entropy alloys, HEA, by two different methods. Both treatments led to the same equation. For cubic HEA lattice parameters were calculated. The predicted lattice parameters were compared with those reported for 68 HEAs. Lattice parameters were also calculated using the equivalent of Vegards law for these alloys. Average errors were 0.52, and 0.42 when Vegards law, and the equation derived in this work were used, respectively.
2104.14541
737,909
We discuss a model where a mixed warm and hot keV neutrino dark matter rises naturally. We arrange active and sterile neutrinos in the same $SU(3)_L$ multiplet, with the lightest sterile neutrino being dark matter. The other two heavy sterile neutrinos, through their out-of-equilibrium decay, contribute both to the dilution of dark matter density and its population, after freeze-out. We show that this model features all ingredients to overcome the overproduction of keV neutrino dark matter, and explore the phenomenological implications for Big Bang Nucleosynthesis and the number of relativistic degrees of freedom.
2104.14542
737,909
Variational quantum algorithms (VQAs) promise efficient use of near-term quantum computers. However, training these algorithms often requires an extensive amount of time and suffers from the barren plateau problem where the magnitude of the gradients vanishes with increasing number of qubits. Here, we show how to optimally train a VQA for learning quantum states. Parameterized quantum circuits can form Gaussian kernels, which we use to derive optimal adaptive learning rates for gradient ascent. We introduce the generalized quantum natural gradient that features stability and optimized movement in parameter space. Both methods together outperform other optimization routines and can enhance VQAs as well as quantum control techniques. The gradients of the VQA do not vanish when the fidelity between the initial state and the state to be learned is bounded from below. We identify a VQA for quantum simulation with such a constraint that can be trained free of barren plateaus. Finally, we propose the application of Gaussian kernels for quantum machine learning.
2104.14543
737,909
Synthetic datasets play a critical role in pre-training CNN models for optical flow, but they are painstaking to generate and hard to adapt to new applications. To automate the process, we present AutoFlow, a simple and effective method to render training data for optical flow that optimizes the performance of a model on a target dataset. AutoFlow takes a layered approach to render synthetic data, where the motion, shape, and appearance of each layer are controlled by learnable hyperparameters. Experimental results show that AutoFlow achieves state-of-the-art accuracy in pre-training both PWC-Net and RAFT. Our code and data are available at https://autoflow-google.github.io .
2104.14544
737,909
Object tracking has achieved significant progress over the past few years. However, state-of-the-art trackers become increasingly heavy and expensive, which limits their deployments in resource-constrained applications. In this work, we present LightTrack, which uses neural architecture search (NAS) to design more lightweight and efficient object trackers. Comprehensive experiments show that our LightTrack is effective. It can find trackers that achieve superior performance compared to handcrafted SOTA trackers, such as SiamRPN++ and Ocean, while using much fewer model Flops and parameters. Moreover, when deployed on resource-constrained mobile chipsets, the discovered trackers run much faster. For example, on Snapdragon 845 Adreno GPU, LightTrack runs $12\times$ faster than Ocean, while using $13\times$ fewer parameters and $38\times$ fewer Flops. Such improvements might narrow the gap between academic models and industrial deployments in object tracking task. LightTrack is released at https://github.com/researchmm/LightTrack.
2104.14545
737,909
Over-parametrized deep neural networks trained by stochastic gradient descent are successful in performing many tasks of practical relevance. One aspect of over-parametrization is the possibility that the student network has a larger expressivity than the data generating process. In the context of a student-teacher scenario, this corresponds to the so-called over-realizable case, where the student network has a larger number of hidden units than the teacher. For on-line learning of a two-layer soft committee machine in the over-realizable case, we find that the approach to perfect learning occurs in a power-law fashion rather than exponentially as in the realizable case. All student nodes learn and replicate one of the teacher nodes if teacher and student outputs are suitably rescaled.
2104.14546
737,909
Recent deep-learning-based techniques for the reconstruction of geometries from different input representations such as images and point clouds have been instrumental in advancing research in geometric machine learning. Most of these techniques rely on a triangular mesh representation for representing the geometry, with very recent attempts in using B-splines. While Non-Uniform Rational B-splines (NURBS) are the de facto standard in the CAD industry, minimal efforts have been made to bridge the gap between deep-learning frameworks and the NURBS representation for geometry. The backbone of modern deep learning techniques is the use of a fully automatic differentiable definition for each mathematical operation to enable backpropagation of losses while training. In order to integrate the NURBS representation of CAD models with deep learning methods, we propose a differentiable NURBS layer for evaluating the curve or surface given a set of NURBS parameters. We have developed a NURBS layer defining the forward and backward pass required for automatic differentiation. Our implementation is GPU accelerated and is directly integrated with PyTorch, a popular deep learning framework. We demonstrate the efficacy of our NURBS layer by automatically incorporating it with the stochastic gradient descent algorithm and performing CAD operations such as curve or surface fitting and surface offsetting. Further, we show its utility in deep learning applications such as point cloud reconstruction and structural modeling and analysis of shell structures such as heart valves. These examples show that our layer has better performance for certain deep learning frameworks and can be directly integrated with any CAD deep-learning framework that require the use of NURBS.
2104.14547
737,909
Self-supervised learning algorithms based on instance discrimination train encoders to be invariant to pre-defined transformations of the same instance. While most methods treat different views of the same image as positives for a contrastive loss, we are interested in using positives from other instances in the dataset. Our method, Nearest-Neighbor Contrastive Learning of visual Representations (NNCLR), samples the nearest neighbors from the dataset in the latent space, and treats them as positives. This provides more semantic variations than pre-defined transformations. We find that using the nearest-neighbor as positive in contrastive losses improves performance significantly on ImageNet classification, from 71.7% to 75.6%, outperforming previous state-of-the-art methods. On semi-supervised learning benchmarks we improve performance significantly when only 1% ImageNet labels are available, from 53.8% to 56.5%. On transfer learning benchmarks our method outperforms state-of-the-art methods (including supervised learning with ImageNet) on 8 out of 12 downstream datasets. Furthermore, we demonstrate empirically that our method is less reliant on complex data augmentations. We see a relative reduction of only 2.1% ImageNet Top-1 accuracy when we train using only random crops.
2104.14548
737,909
This paper proposes a distributed Reinforcement Learning (RL) based framework that can be used for synthesizing MAC layer wireless protocols in IoT networks with low-complexity wireless transceivers. The proposed framework does not rely on complex hardware capabilities such as carrier sensing and its associated algorithmic complexities that are often not supported in wireless transceivers of low-cost and low-energy IoT devices. In this framework, the access protocols are first formulated as Markov Decision Processes (MDP) and then solved using RL. A distributed and multi-Agent RL framework is used as the basis for protocol synthesis. Distributed behavior makes the nodes independently learn optimal transmission strategies without having to rely on full network level information and direct knowledge of behavior of other nodes. The nodes learn to minimize packet collisions such that optimal throughput can be attained and maintained for loading conditions that are higher than what the known benchmark protocols (such as ALOHA) for IoT devices without complex transceivers. In addition, the nodes are observed to be able to learn to act optimally in the presence of heterogeneous loading and network topological conditions. Finally, the proposed learning approach allows the wireless bandwidth to be fairly distributed among network nodes in a way that is not dependent on such heterogeneities. Via simulation experiments, the paper demonstrates the performance of the learning paradigm and its abilities to make nodes adapt their optimal transmission strategies on the fly in response to various network dynamics.
2104.14549
737,909
We show that for a model complete strongly minimal theory whose pregeometry is flat, the recursive spectrum (SRM($T$)) is either of the form $[0,\alpha)$ for $\alpha\in \omega+2$ or $[0,n]\cup\{\omega\}$ for $n\in \omega$, or $\{\omega\}$, or contained in $\{0,1,2\}$. Combined with previous results, this leaves precisely 4 sets for which it is not yet determined whether each is the spectrum of a model complete strongly minimal theory with a flat pregeometry.
2104.14550
737,909
Recent generative models can synthesize "views" of artificial images that mimic real-world variations, such as changes in color or pose, simply by learning from unlabeled image collections. Here, we investigate whether such views can be applied to real images to benefit downstream analysis tasks such as image classification. Using a pretrained generator, we first find the latent code corresponding to a given real input image. Applying perturbations to the code creates natural variations of the image, which can then be ensembled together at test-time. We use StyleGAN2 as the source of generative augmentations and investigate this setup on classification tasks involving facial attributes, cat faces, and cars. Critically, we find that several design decisions are required towards making this process work; the perturbation procedure, weighting between the augmentations and original image, and training the classifier on synthesized images can all impact the result. Currently, we find that while test-time ensembling with GAN-based augmentations can offer some small improvements, the remaining bottlenecks are the efficiency and accuracy of the GAN reconstructions, coupled with classifier sensitivities to artifacts in GAN-generated images.
2104.14551
737,909
Quantum computing has the potential to revolutionize computing for certain classes of problems with exponential scaling, and yet this potential is accompanied by significant sensitivity to noise, requiring sophisticated error correction and mitigation strategies. Here we simulate the relaxations of stationary states at different frequencies on several quantum computers to obtain unique spectroscopic fingerprints of their noise. Response functions generated from the data reveal a clear signature of non-Markovian dynamics, demonstrating that each of the quantum computers acts as a non-Markovian bath with a unique colored noise profile. The study suggest that noisy intermediate-scale quantum computers (NISQ) provide a built-in noisy bath that can be analyzed from their simulation of closed quantum systems with the results potentially being harnessed for error mitigation or open-system simulation.
2104.14552
737,909
Visual content often contains recurring elements. Text is made up of glyphs from the same font, animations, such as cartoons or video games, are composed of sprites moving around the screen, and natural videos frequently have repeated views of objects. In this paper, we propose a deep learning approach for obtaining a graphically disentangled representation of recurring elements in a completely self-supervised manner. By jointly learning a dictionary of texture patches and training a network that places them onto a canvas, we effectively deconstruct sprite-based content into a sparse, consistent, and interpretable representation that can be easily used in downstream tasks. Our framework offers a promising approach for discovering recurring patterns in image collections without supervision.
2104.14553
737,909
Recent advances in geometric deep-learning introduce complex computational challenges for evaluating the distance between meshes. From a mesh model, point clouds are necessary along with a robust distance metric to assess surface quality or as part of the loss function for training models. Current methods often rely on a uniform random mesh discretization, which yields irregular sampling and noisy distance estimation. In this paper we introduce MongeNet, a fast and optimal transport based sampler that allows for an accurate discretization of a mesh with better approximation properties. We compare our method to the ubiquitous random uniform sampling and show that the approximation error is almost half with a very small computational overhead.
2104.14554
737,909
We show how one may classify all semisimple algebras containing the $\mathfrak{su}(3)\oplus \mathfrak{su}(2) \oplus \mathfrak{u}(1)$ symmetry of the Standard Model and acting on some given matter sector, enabling theories beyond the Standard Model with unification (partial or total) of symmetries (gauged or global) to be catalogued. With just a single generation of Standard Model fermions plus a singlet neutrino, the only gauged symmetries correspond to the well-known algebras $\mathfrak{su}(5)$, $\mathfrak{so}(10),$ and $\mathfrak{su}(4)\oplus \mathfrak{su}(2) \oplus \mathfrak{su}(2)$, but with two or more generations a limited number of exotic symmetries mixing flavor, color, and electroweak symmetries become possible. We provide a complete catalogue in the case of 3 generations or fewer and describe how the method can be generalized to include additional matter.
2104.14555
737,909
Recent works find that AI algorithms learn biases from data. Therefore, it is urgent and vital to identify biases in AI algorithms. However, the previous bias identification pipeline overly relies on human experts to conjecture potential biases (e.g., gender), which may neglect other underlying biases not realized by humans. To help human experts better find the AI algorithms' biases, we study a new problem in this work -- for a classifier that predicts a target attribute of the input image, discover its unknown biased attribute. To solve this challenging problem, we use a hyperplane in the generative model's latent space to represent an image attribute; thus, the original problem is transformed to optimizing the hyperplane's normal vector and offset. We propose a novel total-variation loss within this framework as the objective function and a new orthogonalization penalty as a constraint. The latter prevents trivial solutions in which the discovered biased attribute is identical with the target or one of the known-biased attributes. Extensive experiments on both disentanglement datasets and real-world datasets show that our method can discover biased attributes and achieve better disentanglement w.r.t. target attributes. Furthermore, the qualitative results show that our method can discover unnoticeable biased attributes for various object and scene classifiers, proving our method's generalizability for detecting biased attributes in diverse domains of images. The code is available at https://git.io/J3kMh.
2104.14556
737,909
We propose a novel approach for few-shot talking-head synthesis. While recent works in neural talking heads have produced promising results, they can still produce images that do not preserve the identity of the subject in source images. We posit this is a result of the entangled representation of each subject in a single latent code that models 3D shape information, identity cues, colors, lighting and even background details. In contrast, we propose to factorize the representation of a subject into its spatial and style components. Our method generates a target frame in two steps. First, it predicts a dense spatial layout for the target image. Second, an image generator utilizes the predicted layout for spatial denormalization and synthesizes the target frame. We experimentally show that this disentangled representation leads to a significant improvement over previous methods, both quantitatively and qualitatively.
2104.14557
737,909
We present a large-scale study on unsupervised spatiotemporal representation learning from videos. With a unified perspective on four recent image-based frameworks, we study a simple objective that can easily generalize all these methods to space-time. Our objective encourages temporally-persistent features in the same video, and in spite of its simplicity, it works surprisingly well across: (i) different unsupervised frameworks, (ii) pre-training datasets, (iii) downstream datasets, and (iv) backbone architectures. We draw a series of intriguing observations from this study, e.g., we discover that encouraging long-spanned persistency can be effective even if the timespan is 60 seconds. In addition to state-of-the-art results in multiple benchmarks, we report a few promising cases in which unsupervised pre-training can outperform its supervised counterpart. Code is made available at https://github.com/facebookresearch/SlowFast
2104.14558
737,909
Exemplar-based portrait stylization is widely attractive and highly desired. Despite recent successes, it remains challenging, especially when considering both texture and geometric styles. In this paper, we present the first framework for one-shot 3D portrait style transfer, which can generate 3D face models with both the geometry exaggerated and the texture stylized while preserving the identity from the original content. It requires only one arbitrary style image instead of a large set of training examples for a particular style, provides geometry and texture outputs that are fully parameterized and disentangled, and enables further graphics applications with the 3D representations. The framework consists of two stages. In the first geometric style transfer stage, we use facial landmark translation to capture the coarse geometry style and guide the deformation of the dense 3D face geometry. In the second texture style transfer stage, we focus on performing style transfer on the canonical texture by adopting a differentiable renderer to optimize the texture in a multi-view framework. Experiments show that our method achieves robustly good results on different artistic styles and outperforms existing methods. We also demonstrate the advantages of our method via various 2D and 3D graphics applications. Project page is https://halfjoe.github.io/projs/3DPS/index.html.
2104.14559
737,909