February 1, 2020

3044 words 15 mins read

Paper Group AWR 197

Towards Analyzing Semantic Robustness of Deep Neural Networks. Neural Network Distiller: A Python Package For DNN Compression Research. Fast and Robust Dynamic Hand Gesture Recognition via Key Frames Extraction and Feature Fusion. Adversarial Out-domain Examples for Generative Models. Multi-Agent Deep Reinforcement Learning for Large-scale Traffic …

Towards Analyzing Semantic Robustness of Deep Neural Networks


Title	Towards Analyzing Semantic Robustness of Deep Neural Networks
Authors	Abdullah Hamdi, Bernard Ghanem
Abstract	Despite the impressive performance of Deep Neural Networks (DNNs) on various vision tasks, they still exhibit erroneous high sensitivity toward semantic primitives (e.g. object pose). We propose a theoretically grounded analysis for DNNs robustness in the semantic space. We qualitatively analyze different DNNs semantic robustness by visualizing the DNN global behavior as semantic maps and observe interesting behavior of some DNNs. Since generating these semantic maps does not scale well with the dimensionality of the semantic space, we develop a bottom-up approach to detect robust regions of DNNs. To achieve this, We formalize the problem of finding robust semantic regions of the network as optimization of integral bounds and develop expressions for update directions of the region bounds. We use our developed formulations to quantitatively evaluate the semantic robustness of different famous network architectures. We show through extensive experimentation that several networks, though trained on the same dataset and while enjoying comparable accuracy, they do not necessarily perform similarly in semantic robustness. For example, InceptionV3 is more accurate despite being less semantically robust than ResNet50. We hope that this tool will serve as the first milestone towards understanding the semantic robustness of DNNs.
Tasks	Adversarial Attack, Autonomous Driving, Image Classification
Published	2019-04-09
URL	https://arxiv.org/abs/1904.04621v2
PDF	https://arxiv.org/pdf/1904.04621v2.pdf
PWC	https://paperswithcode.com/paper/towards-analyzing-semantic-robustness-of-deep
Repo	https://github.com/ajhamdi/semantic-robustness
Framework	pytorch

Neural Network Distiller: A Python Package For DNN Compression Research


Title	Neural Network Distiller: A Python Package For DNN Compression Research
Authors	Neta Zmora, Guy Jacob, Lev Zlotnik, Bar Elharar, Gal Novik
Abstract	This paper presents the philosophy, design and feature-set of Neural Network Distiller, an open-source Python package for DNN compression research. Distiller is a library of DNN compression algorithms implementations, with tools, tutorials and sample applications for various learning tasks. Its target users are both engineers and researchers, and the rich content is complemented by a design-for-extensibility to facilitate new research. Distiller is open-source and is available on Github at https://github.com/NervanaSystems/distiller.
Tasks
Published	2019-10-27
URL	https://arxiv.org/abs/1910.12232v1
PDF	https://arxiv.org/pdf/1910.12232v1.pdf
PWC	https://paperswithcode.com/paper/neural-network-distiller-a-python-package-for
Repo	https://github.com/NervanaSystems/distiller
Framework	pytorch

Fast and Robust Dynamic Hand Gesture Recognition via Key Frames Extraction and Feature Fusion


Title	Fast and Robust Dynamic Hand Gesture Recognition via Key Frames Extraction and Feature Fusion
Authors	Hao Tang, Hong Liu, Wei Xiao, Nicu Sebe
Abstract	Gesture recognition is a hot topic in computer vision and pattern recognition, which plays a vitally important role in natural human-computer interface. Although great progress has been made recently, fast and robust hand gesture recognition remains an open problem, since the existing methods have not well balanced the performance and the efficiency simultaneously. To bridge it, this work combines image entropy and density clustering to exploit the key frames from hand gesture video for further feature extraction, which can improve the efficiency of recognition. Moreover, a feature fusion strategy is also proposed to further improve feature representation, which elevates the performance of recognition. To validate our approach in a “wild” environment, we also introduce two new datasets called HandGesture and Action3D datasets. Experiments consistently demonstrate that our strategy achieves competitive results on Northwestern University, Cambridge, HandGesture and Action3D hand gesture datasets. Our code and datasets will release at https://github.com/Ha0Tang/HandGestureRecognition.
Tasks	Gesture Recognition, Hand Gesture Recognition, Hand-Gesture Recognition
Published	2019-01-15
URL	http://arxiv.org/abs/1901.04622v1
PDF	http://arxiv.org/pdf/1901.04622v1.pdf
PWC	https://paperswithcode.com/paper/fast-and-robust-dynamic-hand-gesture
Repo	https://github.com/Ha0Tang/HandGestureRecognition
Framework	none

Adversarial Out-domain Examples for Generative Models


Title	Adversarial Out-domain Examples for Generative Models
Authors	Dario Pasquini, Marco Mingione, Massimo Bernaschi
Abstract	Deep generative models are rapidly becoming a common tool for researchers and developers. However, as exhaustively shown for the family of discriminative models, the test-time inference of deep neural networks cannot be fully controlled and erroneous behaviors can be induced by an attacker. In the present work, we show how a malicious user can force a pre-trained generator to reproduce arbitrary data instances by feeding it suitable adversarial inputs. Moreover, we show that these adversarial latent vectors can be shaped so as to be statistically indistinguishable from the set of genuine inputs. The proposed attack technique is evaluated with respect to various GAN images generators using different architectures, training processes and for both conditional and not-conditional setups.
Tasks	Adversarial Attack, Image Generation
Published	2019-03-07
URL	https://arxiv.org/abs/1903.02926v2
PDF	https://arxiv.org/pdf/1903.02926v2.pdf
PWC	https://paperswithcode.com/paper/out-domain-examples-for-generative-models
Repo	https://github.com/pasquini-dario/OutDomainExamples
Framework	tf

Multi-Agent Deep Reinforcement Learning for Large-scale Traffic Signal Control


Title	Multi-Agent Deep Reinforcement Learning for Large-scale Traffic Signal Control
Authors	Tianshu Chu, Jie Wang, Lara Codecà, Zhaojian Li
Abstract	Reinforcement learning (RL) is a promising data-driven approach for adaptive traffic signal control (ATSC) in complex urban traffic networks, and deep neural networks further enhance its learning power. However, centralized RL is infeasible for large-scale ATSC due to the extremely high dimension of the joint action space. Multi-agent RL (MARL) overcomes the scalability issue by distributing the global control to each local RL agent, but it introduces new challenges: now the environment becomes partially observable from the viewpoint of each local agent due to limited communication among agents. Most existing studies in MARL focus on designing efficient communication and coordination among traditional Q-learning agents. This paper presents, for the first time, a fully scalable and decentralized MARL algorithm for the state-of-the-art deep RL agent: advantage actor critic (A2C), within the context of ATSC. In particular, two methods are proposed to stabilize the learning procedure, by improving the observability and reducing the learning difficulty of each local agent. The proposed multi-agent A2C is compared against independent A2C and independent Q-learning algorithms, in both a large synthetic traffic grid and a large real-world traffic network of Monaco city, under simulated peak-hour traffic dynamics. Results demonstrate its optimality, robustness, and sample efficiency over other state-of-the-art decentralized MARL algorithms.
Tasks	Q-Learning
Published	2019-03-11
URL	http://arxiv.org/abs/1903.04527v1
PDF	http://arxiv.org/pdf/1903.04527v1.pdf
PWC	https://paperswithcode.com/paper/multi-agent-deep-reinforcement-learning-for-2
Repo	https://github.com/cts198859/deeprl_signal_control
Framework	tf

Privacy-preserving Q-Learning with Functional Noise in Continuous State Spaces


Title	Privacy-preserving Q-Learning with Functional Noise in Continuous State Spaces
Authors	Baoxiang Wang, Nidhi Hegde
Abstract	We consider differentially private algorithms for reinforcement learning in continuous spaces, such that neighboring reward functions are indistinguishable. This protects the reward information from being exploited by methods such as inverse reinforcement learning. Existing studies that guarantee differential privacy are not extendable to infinite state spaces, as the noise level to ensure privacy will scale accordingly to infinity. Our aim is to protect the value function approximator, without regard to the number of states queried to the function. It is achieved by adding functional noise to the value function iteratively in the training. We show rigorous privacy guarantees by a series of analyses on the kernel of the noise space, the probabilistic bound of such noise samples, and the composition over the iterations. We gain insight into the utility analysis by proving the algorithm’s approximate optimality when the state space is discrete. Experiments corroborate our theoretical findings and show improvement over existing approaches.
Tasks	Q-Learning
Published	2019-01-30
URL	https://arxiv.org/abs/1901.10634v3
PDF	https://arxiv.org/pdf/1901.10634v3.pdf
PWC	https://paperswithcode.com/paper/private-q-learning-with-functional-noise-in
Repo	https://github.com/wangbx66/differentially-private-q-learning
Framework	pytorch

Multi-Hop Paragraph Retrieval for Open-Domain Question Answering


Title	Multi-Hop Paragraph Retrieval for Open-Domain Question Answering
Authors	Yair Feldman, Ran El-Yaniv
Abstract	This paper is concerned with the task of multi-hop open-domain Question Answering (QA). This task is particularly challenging since it requires the simultaneous performance of textual reasoning and efficient searching. We present a method for retrieving multiple supporting paragraphs, nested amidst a large knowledge base, which contain the necessary evidence to answer a given question. Our method iteratively retrieves supporting paragraphs by forming a joint vector representation of both a question and a paragraph. The retrieval is performed by considering contextualized sentence-level representations of the paragraphs in the knowledge source. Our method achieves state-of-the-art performance over two well-known datasets, SQuAD-Open and HotpotQA, which serve as our single- and multi-hop open-domain QA benchmarks, respectively.
Tasks	Open-Domain Question Answering, Question Answering
Published	2019-06-15
URL	https://arxiv.org/abs/1906.06606v1
PDF	https://arxiv.org/pdf/1906.06606v1.pdf
PWC	https://paperswithcode.com/paper/multi-hop-paragraph-retrieval-for-open-domain
Repo	https://github.com/yairf11/MUPPET
Framework	tf


Title	The Regretful Agent: Heuristic-Aided Navigation through Progress Estimation
Authors	Chih-Yao Ma, Zuxuan Wu, Ghassan AlRegib, Caiming Xiong, Zsolt Kira
Abstract	As deep learning continues to make progress for challenging perception tasks, there is increased interest in combining vision, language, and decision-making. Specifically, the Vision and Language Navigation (VLN) task involves navigating to a goal purely from language instructions and visual information without explicit knowledge of the goal. Recent successful approaches have made in-roads in achieving good success rates for this task but rely on beam search, which thoroughly explores a large number of trajectories and is unrealistic for applications such as robotics. In this paper, inspired by the intuition of viewing the problem as search on a navigation graph, we propose to use a progress monitor developed in prior work as a learnable heuristic for search. We then propose two modules incorporated into an end-to-end architecture: 1) A learned mechanism to perform backtracking, which decides whether to continue moving forward or roll back to a previous state (Regret Module) and 2) A mechanism to help the agent decide which direction to go next by showing directions that are visited and their associated progress estimate (Progress Marker). Combined, the proposed approach significantly outperforms current state-of-the-art methods using greedy action selection, with 5% absolute improvement on the test server in success rates, and more importantly 8% on success rates normalized by the path length. Our code is available at https://github.com/chihyaoma/regretful-agent .
Tasks	Decision Making, Vision-Language Navigation, Visual Navigation
Published	2019-03-05
URL	http://arxiv.org/abs/1903.01602v1
PDF	http://arxiv.org/pdf/1903.01602v1.pdf
PWC	https://paperswithcode.com/paper/the-regretful-agent-heuristic-aided
Repo	https://github.com/chihyaoma/regretful-agent
Framework	pytorch

Marathon Environments: Multi-Agent Continuous Control Benchmarks in a Modern Video Game Engine


Title	Marathon Environments: Multi-Agent Continuous Control Benchmarks in a Modern Video Game Engine
Authors	Joe Booth, Jackson Booth
Abstract	Recent advances in deep reinforcement learning in the paradigm of locomotion using continuous control have raised the interest of game makers for the potential of digital actors using active ragdoll. Currently, the available options to develop these ideas are either researchers’ limited codebase or proprietary closed systems. We present Marathon Environments, a suite of open source, continuous control benchmarks implemented on the Unity game engine, using the Unity ML- Agents Toolkit. We demonstrate through these benchmarks that continuous control research is transferable to a commercial game engine. Furthermore, we exhibit the robustness of these environments by reproducing advanced continuous control research, such as learning to walk, run and backflip from motion capture data; learning to navigate complex terrains; and by implementing a video game input control system. We show further robustness by training with alternative algorithms found in OpenAI.Baselines. Finally, we share strategies for significantly reducing the training time.
Tasks	Continuous Control, Motion Capture
Published	2019-02-25
URL	http://arxiv.org/abs/1902.09097v1
PDF	http://arxiv.org/pdf/1902.09097v1.pdf
PWC	https://paperswithcode.com/paper/marathon-environments-multi-agent-continuous
Repo	https://github.com/Unity-Technologies/marathon-envs
Framework	none

DIANet: Dense-and-Implicit Attention Network


Title	DIANet: Dense-and-Implicit Attention Network
Authors	Zhongzhan Huang, Senwei Liang, Mingfu Liang, Haizhao Yang
Abstract	Attention networks have successfully boosted the performance in various vision problems. Previous works lay emphasis on designing a new attention module and individually plug them into the networks. Our paper proposes a novel-and-simple framework that shares an attention module throughout different network layers to encourage the integration of layer-wise information and this parameter-sharing module is referred as Dense-and-Implicit-Attention (DIA) unit. Many choices of modules can be used in the DIA unit. Since Long Short Term Memory (LSTM) has a capacity of capturing long-distance dependency, we focus on the case when the DIA unit is the modified LSTM (refer as DIA-LSTM). Experiments on benchmark datasets show that the DIA-LSTM unit is capable of emphasizing layer-wise feature interrelation and leads to significant improvement of image classification accuracy. We further empirically show that the DIA-LSTM has a strong regularization ability on stabilizing the training of deep networks by the experiments with the removal of skip connections or Batch Normalization in the whole residual network. The code is released at https://github.com/gbup-group/DIANet.
Tasks	Image Classification
Published	2019-05-25
URL	https://arxiv.org/abs/1905.10671v2
PDF	https://arxiv.org/pdf/1905.10671v2.pdf
PWC	https://paperswithcode.com/paper/dianet-dense-and-implicit-attention-network
Repo	https://github.com/osmr/imgclsmob
Framework	mxnet

Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluation


Title	Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluation
Authors	Ruibo Tu, Kun Zhang, Bo Christer Bertilson, Hedvig Kjellström, Cheng Zhang
Abstract	Discovery of causal relations from observational data is essential for many disciplines of science and real-world applications. However, unlike other machine learning algorithms, whose development has been greatly fostered by a large amount of available benchmark datasets, causal discovery algorithms are notoriously difficult to be systematically evaluated because few datasets with known ground-truth causal relations are available. In this work, we handle the problem of evaluating causal discovery algorithms by building a flexible simulator in the medical setting. We develop a neuropathic pain diagnosis simulator, inspired by the fact that the biological processes of neuropathic pathophysiology are well studied with well-understood causal influences. Our simulator exploits the causal graph of the neuropathic pain pathology and its parameters in the generator are estimated from real-life patient cases. We show that the data generated from our simulator have similar statistics as real-world data. As a clear advantage, the simulator can produce infinite samples without jeopardizing the privacy of real-world patients. Our simulator provides a natural tool for evaluating various types of causal discovery algorithms, including those to deal with practical issues in causal discovery, such as unknown confounders, selection bias, and missing data. Using our simulator, we have evaluated extensively causal discovery algorithms under various settings.
Tasks	Causal Discovery
Published	2019-06-04
URL	https://arxiv.org/abs/1906.01732v3
PDF	https://arxiv.org/pdf/1906.01732v3.pdf
PWC	https://paperswithcode.com/paper/neuropathic-pain-diagnosis-simulator-for
Repo	https://github.com/TURuibo/Neuropathic-Pain-Diagnosis-Simulator
Framework	none

Emergence of Exploratory Look-Around Behaviors through Active Observation Completion


Title	Emergence of Exploratory Look-Around Behaviors through Active Observation Completion
Authors	Santhosh K. Ramakrishnan, Dinesh Jayaraman, Kristen Grauman
Abstract	Standard computer vision systems assume access to intelligently captured inputs (e.g., photos from a human photographer), yet autonomously capturing good observations is a major challenge in itself. We address the problem of learning to look around: how can an agent learn to acquire informative visual observations? We propose a reinforcement learning solution, where the agent is rewarded for reducing its uncertainty about the unobserved portions of its environment. Specifically, the agent is trained to select a short sequence of glimpses after which it must infer the appearance of its full environment. To address the challenge of sparse rewards, we further introduce sidekick policy learning, which exploits the asymmetry in observability between training and test time. The proposed methods learn observation policies that not only perform the completion task for which they are trained, but also generalize to exhibit useful “look-around” behavior for a range of active perception tasks.
Tasks	Active Observation Completion, Observation Completion
Published	2019-06-27
URL	https://arxiv.org/abs/1906.11407v1
PDF	https://arxiv.org/pdf/1906.11407v1.pdf
PWC	https://paperswithcode.com/paper/emergence-of-exploratory-look-around-1
Repo	https://github.com/srama2512/sidekicks
Framework	pytorch

A Debiased MDI Feature Importance Measure for Random Forests


Title	A Debiased MDI Feature Importance Measure for Random Forests
Authors	Xiao Li, Yu Wang, Sumanta Basu, Karl Kumbier, Bin Yu
Abstract	Tree ensembles such as Random Forests have achieved impressive empirical success across a wide variety of applications. To understand how these models make predictions, people routinely turn to feature importance measures calculated from tree ensembles. It has long been known that Mean Decrease Impurity (MDI), one of the most widely used measures of feature importance, incorrectly assigns high importance to noisy features, leading to systematic bias in feature selection. In this paper, we address the feature selection bias of MDI from both theoretical and methodological perspectives. Based on the original definition of MDI by Breiman et al. for a single tree, we derive a tight non-asymptotic bound on the expected bias of MDI importance of noisy features, showing that deep trees have higher (expected) feature selection bias than shallow ones. However, it is not clear how to reduce the bias of MDI using its existing analytical expression. We derive a new analytical expression for MDI, and based on this new expression, we are able to propose a debiased MDI feature importance measure using out-of-bag samples, called MDI-oob. For both the simulated data and a genomic ChIP dataset, MDI-oob achieves state-of-the-art performance in feature selection from Random Forests for both deep and shallow trees.
Tasks	Feature Importance, Feature Selection
Published	2019-06-26
URL	https://arxiv.org/abs/1906.10845v2
PDF	https://arxiv.org/pdf/1906.10845v2.pdf
PWC	https://paperswithcode.com/paper/a-debiased-mdi-feature-importance-measure-for
Repo	https://github.com/nalzok/tree.interpreter
Framework	none

Asking Easy Questions: A User-Friendly Approach to Active Reward Learning


Title	Asking Easy Questions: A User-Friendly Approach to Active Reward Learning
Authors	Erdem Bıyık, Malayandi Palan, Nicholas C. Landolfi, Dylan P. Losey, Dorsa Sadigh
Abstract	Robots can learn the right reward function by querying a human expert. Existing approaches attempt to choose questions where the robot is most uncertain about the human’s response; however, they do not consider how easy it will be for the human to answer! In this paper we explore an information gain formulation for optimally selecting questions that naturally account for the human’s ability to answer. Our approach identifies questions that optimize the trade-off between robot and human uncertainty, and determines when these questions become redundant or costly. Simulations and a user study show our method not only produces easy questions, but also ultimately results in faster reward learning.
Tasks
Published	2019-10-10
URL	https://arxiv.org/abs/1910.04365v1
PDF	https://arxiv.org/pdf/1910.04365v1.pdf
PWC	https://paperswithcode.com/paper/asking-easy-questions-a-user-friendly
Repo	https://github.com/Stanford-ILIAD/easy-active-learning
Framework	none

Multimodal Image Outpainting With Regularized Normalized Diversification


Title	Multimodal Image Outpainting With Regularized Normalized Diversification
Authors	Lingzhi Zhang, Jiancong Wang, Jianbo Shi
Abstract	In this paper, we study the problem of generating a set ofrealistic and diverse backgrounds when given only a smallforeground region. We refer to this task as image outpaint-ing. The technical challenge of this task is to synthesize notonly plausible but also diverse image outputs. Traditionalgenerative adversarial networks suffer from mode collapse.While recent approaches propose to maximize orpreserve the pairwise distance between generated sampleswith respect to their latent distance, they do not explicitlyprevent the diverse samples of different conditional inputsfrom collapsing. Therefore, we propose a new regulariza-tion method to encourage diverse sampling in conditionalsynthesis. In addition, we propose a feature pyramid dis-criminator to improve the image quality. Our experimen-tal results show that our model can produce more diverseimages without sacrificing visual quality compared to state-of-the-arts approaches in both the CelebA face dataset and the Cityscape scene dataset.
Tasks	Image Outpainting
Published	2019-10-25
URL	https://arxiv.org/abs/1910.11481v1
PDF	https://arxiv.org/pdf/1910.11481v1.pdf
PWC	https://paperswithcode.com/paper/multimodal-image-outpainting-with-regularized
Repo	https://github.com/owenzlz/Diverse_Outpaint
Framework	pytorch