Paper Group NANR 84
HOPPITY: LEARNING GRAPH TRANSFORMATIONS TO DETECT AND FIX BUGS IN PROGRAMS. Benchmarking Model-Based Reinforcement Learning. Defending Against Physically Realizable Attacks on Image Classification. 3D Human Pose Estimation using Spatio-Temporal Networks with Explicit Occlusion Training. The Gambler’s Problem and Beyond. Smooth markets: A basic mech …
HOPPITY: LEARNING GRAPH TRANSFORMATIONS TO DETECT AND FIX BUGS IN PROGRAMS
| Title | HOPPITY: LEARNING GRAPH TRANSFORMATIONS TO DETECT AND FIX BUGS IN PROGRAMS |
| Authors | Anonymous |
| Abstract | We present a learning-based approach to detect and fix a broad range of bugs in Javascript programs. We frame the problem in terms of learning a sequence of graph transformations: given a buggy program modeled by a graph structure, our model makes a sequence of predictions including the position of bug nodes and corresponding graph edits to produce a fix. Unlike previous works that use deep neural networks, our approach targets bugs that are more complex and semantic in nature (i.e.~bugs that require adding or deleting statements to fix). We have realized our approach in a tool called HOPPITY. By training on 338,877 Javascript code change commits on Github, HOPPITY correctly detects and fixes bugs in 9,612 out of 42,365 programs in an end-to-end fashion. Given the bug location and type of the fix, HOPPITY also outperforms the baseline approach by a wide margin. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=SJeqs6EFvB |
| https://openreview.net/pdf?id=SJeqs6EFvB | |
| PWC | https://paperswithcode.com/paper/hoppity-learning-graph-transformations-to |
| Repo | |
| Framework | |
Benchmarking Model-Based Reinforcement Learning
| Title | Benchmarking Model-Based Reinforcement Learning |
| Authors | Anonymous |
| Abstract | Model-based reinforcement learning (MBRL) is widely seen as having the potential to be significantly more sample efficient than model-free RL. However, research in model-based RL has not been very standardized. It is fairly common for authors to experiment with self-designed environments, and there are several separate lines of research, which are sometimes closed-sourced or not reproducible. Accordingly, it is an open question how these various existing algorithms perform relative to each other. To facilitate research in MBRL, in this paper we gather a wide collection of MBRL algorithms and propose over 18 benchmarking environments specially designed for MBRL. We benchmark these algorithms with unified problem settings, including noisy environments. Beyond cataloguing performance, we explore and unify the underlying algorithmic differences across MBRL algorithms. We characterize three key research challenges for future MBRL research: the dynamics bottleneck, the planning horizon dilemma, and the early-termination dilemma. Finally, to facilitate future research on MBRL, we open-source our benchmark. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=H1lefTEKDS |
| https://openreview.net/pdf?id=H1lefTEKDS | |
| PWC | https://paperswithcode.com/paper/benchmarking-model-based-reinforcement-1 |
| Repo | |
| Framework | |
Defending Against Physically Realizable Attacks on Image Classification
| Title | Defending Against Physically Realizable Attacks on Image Classification |
| Authors | Anonymous |
| Abstract | We study the problem of defending deep neural network approaches for image classification from physically realizable attacks. First, we demonstrate that the two most scalable and effective methods for learning robust models, adversarial training with PGD attacks and randomized smoothing, exhibit very limited effectiveness against three of the highest profile physical attacks. Next, we propose a new abstract adversarial model, rectangular occlusion attacks, in which an adversary places a small adversarially crafted rectangle in an image, and develop two approaches for efficiently computing the resulting adversarial examples. Finally, we demonstrate that adversarial training using our new attack yields image classification models that exhibit high robustness against the physically realizable attacks we study, offering the first effective generic defense against such attacks. |
| Tasks | Image Classification |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=H1xscnEKDr |
| https://openreview.net/pdf?id=H1xscnEKDr | |
| PWC | https://paperswithcode.com/paper/defending-against-physically-realizable-1 |
| Repo | |
| Framework | |
3D Human Pose Estimation using Spatio-Temporal Networks with Explicit Occlusion Training
| Title | 3D Human Pose Estimation using Spatio-Temporal Networks with Explicit Occlusion Training |
| Authors | Cheng Yu, Bo Yang, Bo Wang, Robby T. Tan |
| Abstract | Estimating 3D poses from a monocular video is still a challenging task, despite the significant progress that has been made in the recent years. Generally, the performance of existing methods drops when the target person is too small/large, or the motion is too fast/slow relative to the scale and speed of the training data. Moreover, to our knowledge, many of these methods are not designed or trained under severe occlusion explicitly, making their performance on handling occlusion compromised. Addressing these problems, we introduce a spatio-temporal network for robust 3D human pose estimation. As humans in videos may appear in different scales and have various motion speeds, we apply multi-scale spatial features for 2D joints or keypoints prediction in each individual frame, and multi-stride temporal convolutional networks (TCNs) to estimate 3D joints or keypoints. Furthermore, we design a spatio-temporal discriminator based on body structures as well as limb motions to assess whether the predicted pose forms a valid pose and a valid movement. During training, we explicitly mask out some keypoints to simulate various occlusion cases, from minor to severe occlusion, so that our network can learn better and becomes robust to various degrees of occlusion. As there are limited 3D ground truth data, we further utilize 2D video data to inject a semi-supervised learning capability to our network. Experiments on public data sets validate the effectiveness of our method, and our ablation studies show the strengths of our network’s individual submodules. |
| Tasks | 3D Human Pose Estimation, Pose Estimation |
| Published | 2020-02-07 |
| URL | https://aaai.org/Conferences/AAAI-20/wp-content/uploads/2020/02/AAAI20-ProgramWeb.pdf |
| https://www.dropbox.com/s/tm8q3agp4chtlnz/3DPoseEstimation_AAAI20.pdf?dl=0 | |
| PWC | https://paperswithcode.com/paper/3d-human-pose-estimation-using-spatio |
| Repo | |
| Framework | |
The Gambler’s Problem and Beyond
| Title | The Gambler’s Problem and Beyond |
| Authors | Anonymous |
| Abstract | We analyze the Gambler’s problem, a simple reinforcement learning problem where the gambler has the chance to double or lose their bets until the target is reached. This is an early example introduced in the reinforcement learning textbook by \cite{sutton2018reinforcement}, where they mention an interesting pattern of the optimal value function with high-frequency components and repeating non-smooth points but without further investigation. We provide the exact formula for the optimal value function for both the discrete and the continuous case. Though simple as it might seem, the value function is pathological: fractal, self-similar, non-smooth on any interval, zero derivative almost everywhere, and not written as elementary functions. Sharing these properties with the Cantor function, it holds a complexity that has been uncharted thus far. With the analysis, our work could lead insights on improving value function approximation, Q-learning, and gradient-based algorithms in real applications and implementations. |
| Tasks | Q-Learning |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=HyxnMyBKwB |
| https://openreview.net/pdf?id=HyxnMyBKwB | |
| PWC | https://paperswithcode.com/paper/the-gamblers-problem-and-beyond |
| Repo | |
| Framework | |
Smooth markets: A basic mechanism for organizing gradient-based learners
| Title | Smooth markets: A basic mechanism for organizing gradient-based learners |
| Authors | Anonymous |
| Abstract | With the success of modern machine learning, it is becoming increasingly important to understand and control how learning algorithms interact. Unfortunately, negative results from game theory show there is little hope of understanding or controlling general n-player games. We therefore introduce smooth markets (SM-games), a class of n-player games with pairwise zero sum interactions. SM-games codify a common design pattern in machine learning that includes some GANs, adversarial training, and other recent algorithms. We show that SM-games are amenable to analysis and optimization using first-order methods. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=B1xMEerYvB |
| https://openreview.net/pdf?id=B1xMEerYvB | |
| PWC | https://paperswithcode.com/paper/smooth-markets-a-basic-mechanism-for |
| Repo | |
| Framework | |
Gating Revisited: Deep Multi-layer RNNs That Can Be Trained
| Title | Gating Revisited: Deep Multi-layer RNNs That Can Be Trained |
| Authors | Anonymous |
| Abstract | Recurrent Neural Networks (RNNs) are widely used models for sequence data. Just like for feedforward networks, it has become common to build “deep” RNNs, i.e., stack multiple recurrent layers to obtain higher-level abstractions of the data. However, this works only for a handful of layers. Unlike feedforward networks, stacking more than a few recurrent units (e.g., LSTM cells) usually hurts model performance, the reason being vanishing or exploding gradients during training. We investigate the training of multi-layer RNNs and examine the magnitude of the gradients as they propagate through the network. We show that, depending on the structure of the basic recurrent unit, the gradients are systematically attenuated or amplified, so that with an increasing depth they tend to vanish, respectively explode. Based on our analysis we design a new type of gated cell that better preserves gradient magnitude, and therefore makes it possible to train deeper RNNs. We experimentally validate our design with five different sequence modelling tasks on three different datasets. The proposed stackable recurrent (STAR) cell allows for substantially deeper recurrent architectures, with improved performance. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=SkgNZeSKPB |
| https://openreview.net/pdf?id=SkgNZeSKPB | |
| PWC | https://paperswithcode.com/paper/gating-revisited-deep-multi-layer-rnns-that |
| Repo | |
| Framework | |
Do recent advancements in model-based deep reinforcement learning really improve data efficiency?
| Title | Do recent advancements in model-based deep reinforcement learning really improve data efficiency? |
| Authors | Anonymous |
| Abstract | Reinforcement learning (RL) has seen great advancements in the past few years. Nevertheless, the consensus among the RL community is that currently used model-free methods, despite all their benefits, suffer from extreme data inefficiency. To circumvent this problem, novel model-based approaches were introduced that often claim to be much more efficient than their model-free counterparts. In this paper, however, we demonstrate that the state-of-the-art model-free Rainbow DQN algorithm can be trained using a much smaller number of samples than it is commonly reported. By simply allowing the algorithm to execute network updates more frequently we manage to reach similar or better results than existing model-based techniques, at a fraction of complexity and computational costs. Furthermore, based on the outcomes of the study, we argue that the agent similar to the modified Rainbow DQN that is presented in this paper should be used as a baseline for any future work aimed at improving sample efficiency of deep reinforcement learning. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=Bke9u1HFwB |
| https://openreview.net/pdf?id=Bke9u1HFwB | |
| PWC | https://paperswithcode.com/paper/do-recent-advancements-in-model-based-deep |
| Repo | |
| Framework | |
Universality Theorems for Generative Models
| Title | Universality Theorems for Generative Models |
| Authors | Anonymous |
| Abstract | Despite the fact that generative models are extremely successful in practice, the theory underlying this phenomenon is only starting to catch up with practice. In this work we address the question of the universality of generative models: is it true that neural networks can approximate any data manifold arbitrarily well? We provide a positive answer to this question and show that under mild assumptions on the activation function one can always find a feedforward neural network that maps the latent space onto a set located within the specified Hausdorff distance from the desired data manifold. We also prove similar theorems for the case of multiclass generative models and cycle generative models, trained to map samples from one manifold to another and vice versa. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=rJlJF1SYPB |
| https://openreview.net/pdf?id=rJlJF1SYPB | |
| PWC | https://paperswithcode.com/paper/universality-theorems-for-generative-models-1 |
| Repo | |
| Framework | |
The Dual Information Bottleneck
| Title | The Dual Information Bottleneck |
| Authors | Anonymous |
| Abstract | The Information-Bottleneck (IB) framework suggests a general characterization of optimal representations in learning, and deep learning in particular. It is based on the optimal trade off between the representation complexity and accuracy, both of which are quantified by mutual information. The problem is solved by alternating projections between the encoder and decoder of the representation, which can be performed locally at each representation level. The framework, however, has practical drawbacks, in that mutual information is notoriously difficult to handle at high dimension, and only has closed form solutions in special cases. Further, because it aims to extract representations which are minimal sufficient statistics of the data with respect to the desired label, it does not necessarily optimize the actual prediction of unseen labels. Here we present a formal dual problem to the IB which has several interesting properties. By switching the order in the KL-divergence between the representation decoder and data, the optimal decoder becomes the geometric rather than the arithmetic mean of the input points. While providing a good approximation to the original IB, it also preserves the form of exponential families, and optimizes the mutual information on the predicted label rather than the desired one. We also analyze the critical points of the dualIB and discuss their importance for the quality of this approach. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=B1xZD1rtPr |
| https://openreview.net/pdf?id=B1xZD1rtPr | |
| PWC | https://paperswithcode.com/paper/the-dual-information-bottleneck |
| Repo | |
| Framework | |
SemanticAdv: Generating Adversarial Examples via Attribute-Conditional Image Editing
| Title | SemanticAdv: Generating Adversarial Examples via Attribute-Conditional Image Editing |
| Authors | Anonymous |
| Abstract | Deep neural networks (DNNs) have achieved great success in various applications due to their strong expressive power. However, recent studies have shown that DNNs are vulnerable to adversarial examples which are manipulated instances targeting to mislead DNNs to make incorrect predictions. Currently, most such adversarial examples try to guarantee “subtle perturbation” by limiting the Lp norm of the perturbation. In this paper, we aim to explore the impact of semantic manipulation on DNNs predictions by manipulating the semantic attributes of images and generate “unrestricted adversarial examples”. Such semantic based perturbation is more practical compared with the Lp bounded perturbation. In particular, we propose an algorithm SemanticAdv which leverages disentangled semantic factors to generate adversarial perturbation by altering controlled semantic attributes to fool the learner towards various “adversarial” targets. We conduct extensive experiments to show that the semantic based adversarial examples can not only fool different learning tasks such as face verification and landmark detection, but also achieve high targeted attack success rate against real-world black-box services such as Azure face verification service based on transferability. To further demonstrate the applicability of SemanticAdv beyond face recognition domain, we also generate semantic perturbations on street-view images. Such adversarial examples with controlled semantic manipulation can shed light on further understanding about vulnerabilities of DNNs as well as potential defensive approaches. |
| Tasks | Face Recognition, Face Verification |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=r1l-VeSKwS |
| https://openreview.net/pdf?id=r1l-VeSKwS | |
| PWC | https://paperswithcode.com/paper/semanticadv-generating-adversarial-examples-1 |
| Repo | |
| Framework | |
Learning to solve the credit assignment problem
| Title | Learning to solve the credit assignment problem |
| Authors | Anonymous |
| Abstract | Backpropagation is driving today’s artificial neural networks (ANNs). However, despite extensive research, it remains unclear if the brain implements this algorithm. Among neuroscientists, reinforcement learning (RL) algorithms are often seen as a realistic alternative: neurons can randomly introduce change, and use unspecific feedback signals to observe their effect on the cost and thus approximate their gradient. However, the convergence rate of such learning scales poorly with the number of involved neurons. Here we propose a hybrid learning approach. Each neuron uses an RL-type strategy to learn how to approximate the gradients that backpropagation would provide. We provide proof that our approach converges to the true gradient for certain classes of networks. In both feedforward and convolutional networks, we empirically show that our approach learns to approximate the gradient, and can match the performance of gradient-based learning. Learning feedback weights provides a biologically plausible mechanism of achieving good performance, without the need for precise, pre-specified learning rules. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=ByeUBANtvB |
| https://openreview.net/pdf?id=ByeUBANtvB | |
| PWC | https://paperswithcode.com/paper/learning-to-solve-the-credit-assignment |
| Repo | |
| Framework | |
Scaleable input gradient regularization for adversarial robustness
| Title | Scaleable input gradient regularization for adversarial robustness |
| Authors | Anonymous |
| Abstract | In this work we revisit gradient regularization for adversarial robustness with some new ingredients. First, we derive new per-image theoretical robustness bounds based on local gradient information. These bounds strongly motivate input gradient regularization. Second, we implement a scaleable version of input gradient regularization which avoids double backpropagation: adversarially robust ImageNet models are trained in 33 hours on four consumer grade GPUs. Finally, we show experimentally and through theoretical certification that input gradient regularization is competitive with adversarial training. Moreover we demonstrate that gradient regularization does not lead to gradient obfuscation or gradient masking. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=Hkl_bCVKDr |
| https://openreview.net/pdf?id=Hkl_bCVKDr | |
| PWC | https://paperswithcode.com/paper/scaleable-input-gradient-regularization-for-1 |
| Repo | |
| Framework | |
Fractional Graph Convolutional Networks (FGCN) for Semi-Supervised Learning
| Title | Fractional Graph Convolutional Networks (FGCN) for Semi-Supervised Learning |
| Authors | Anonymous |
| Abstract | Due to high utility in many applications, from social networks to blockchain to power grids, deep learning on non-Euclidean objects such as graphs and manifolds continues to gain an ever increasing interest. Most currently available techniques are based on the idea of performing a convolution operation in the spectral domain with a suitably chosen nonlinear trainable filter and then approximating the filter with finite order polynomials. However, such polynomial approximation approaches tend to be both non-robust to changes in the graph structure and to capture primarily the global graph topology. In this paper we propose a new Fractional Generalized Graph Convolutional Networks (FGCN) method for semi-supervised learning, which casts the L'evy Fights into random walks on graphs and, as a result, allows to more accurately account for the intrinsic graph topology and to substantially improve classification performance, especially for heterogeneous graphs. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=BygacxrFwS |
| https://openreview.net/pdf?id=BygacxrFwS | |
| PWC | https://paperswithcode.com/paper/fractional-graph-convolutional-networks-fgcn |
| Repo | |
| Framework | |
Insights on Visual Representations for Embodied Navigation Tasks
| Title | Insights on Visual Representations for Embodied Navigation Tasks |
| Authors | Erik Wijmans, Julian Straub, Irfan Essa, Dhruv Batra, Judy Hoffman, Ari Morcos |
| Abstract | Recent advances in deep reinforcement learning require a large amount of training data and generally result in representations that are often over specialized to the target task. In this work, we study the underlying potential causes for this specialization by measuring the similarity between representations trained on related, but distinct tasks. We use the recently proposed projection weighted Canonical Correlation Analysis (PWCCA) to examine the task dependence of visual representations learned across different embodied navigation tasks. Surprisingly, we find that slight differences in task have no measurable effect on the visual representation for both SqueezeNet and ResNet architectures. We then empirically demonstrate that visual representations learned on one task can be effectively transferred to a different task. Interestingly, we show that if the tasks constrain the agent to spatially disjoint parts of the environment, differences in representation emerge for SqueezeNet models but less-so for ResNets, suggesting that ResNets feature inductive biases which encourage more task-agnostic representations, even in the context of spatially separated tasks. We generalize our analysis to examine permutations of an environment and find, surprisingly, permutations of an environment also do not influence the visual representation. Our analysis provides insight on the overfitting of representations in RL and provides suggestions of how to design tasks that induce task-agnostic representations. |
| Tasks | |
| Published | 2020-01-01 |
| URL | https://openreview.net/forum?id=BkxadR4KvS |
| https://openreview.net/pdf?id=BkxadR4KvS | |
| PWC | https://paperswithcode.com/paper/insights-on-visual-representations-for |
| Repo | |
| Framework | |
