2020.06.15.txt

==========New Papers==========
1, TITLE:       A Brief Look at Generalization in Visual Meta-Reinforcement Learning
http://arxiv.org/abs/2006.07262
AUTHORS:        Safa Alver ; Doina Precup
COMMENTS:       8 pages, 4 figures
HIGHLIGHT:      In this paper, we assess the generalization performance of these algorithms by leveraging high-dimensional, procedurally generated environments.

2, TITLE:       Rethinking Sampling in 3D Point Cloud Generative Adversarial Networks
http://arxiv.org/abs/2006.07029
AUTHORS:        He Wang ; Zetian Jiang ; Li Yi ; Kaichun Mo ; Hao Su ; Leonidas J. Guibas
HIGHLIGHT:      In this paper, we examine the long-neglected yet important effects of point sampling patterns in point cloud GANs.

3, TITLE:       Low-resource Languages: A Review of Past Work and Future Challenges
http://arxiv.org/abs/2006.07264
AUTHORS:        Alexandre Magueresse ; Vincent Carles ; Evan Heetderks
HIGHLIGHT:      This review paper concisely summarizes previous groundbreaking achievements made towards resolving this problem, and analyzes potential improvements in the context of the overall future research direction.

4, TITLE:       Improved Fixed-Budget Results via Drift Analysis
http://arxiv.org/abs/2006.07019
AUTHORS:        Timo Kötzing ; Carsten Witt
COMMENTS:       25 pages. An extended abstract of this paper will be published in the proceedings of PPSN 2020
HIGHLIGHT:      Fixed-budget theory is concerned with computing or bounding the fitness value achievable by randomized search heuristics within a given budget of fitness function evaluations.

5, TITLE:       Learning Effective Representations for Person-Job Fit by Feature Fusion
http://arxiv.org/abs/2006.07017
AUTHORS:        Junshu Jiang ; Songyun Ye ; Wei Wang ; Jingran Xu ; Xiaosheng Luo
COMMENTS:       8 pages
HIGHLIGHT:      In this paper, we propose to learn comprehensive and effective representations of the candidates and job posts via feature fusion.

6, TITLE:       Speaker Sensitive Response Evaluation Model
http://arxiv.org/abs/2006.07015
AUTHORS:        JinYeong Bak ; Alice Oh
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      In this paper, we propose an automatic evaluation model based on that idea and learn the model parameters from an unlabeled conversation corpus.

7, TITLE:       Background Modeling via Uncertainty Estimation for Weakly-supervised Action Localization
http://arxiv.org/abs/2006.07006
AUTHORS:        Pilhyeon Lee ; Jinglu Wang ; Yan Lu ; Hyeran Byun
HIGHLIGHT:      Beyond our base action localization network, we propose a module to estimate the probability of being background (i.e., uncertainty [20]), which allows us to learn uncertainty given only video-level labels via multiple instance learning.

8, TITLE:       Training spiking multi-layer networks with surrogate gradients on an analog neuromorphic substrate
http://arxiv.org/abs/2006.07239
AUTHORS:        Benjamin Cramer ; Sebastian Billaudelle ; Simeon Kanya ; Aron Leibfried ; Andreas Grübl ; Vitali Karasenko ; Christian Pehle ; Korbinian Schreiber ; Yannik Stradmann ; Johannes Weis ; Johannes Schemmel ; Friedemann Zenke
HIGHLIGHT:      We evaluated our approach on downscaled 16x16 versions of the MNIST and the fashion MNIST datasets in which spike latencies encoded pixel intensities.

9, TITLE:       Power Consumption Variation over Activation Functions
http://arxiv.org/abs/2006.07237
AUTHORS:        Leon Derczynski
HIGHLIGHT:      This paper presents various estimates of power consumption for a range of different activation functions, a core factor in neural network model architecture design.

10, TITLE:      SemEval-2020 Task 12: Multilingual Offensive Language Identification in Social Media (OffensEval 2020)
http://arxiv.org/abs/2006.07235
AUTHORS:        Marcos Zampieri ; Preslav Nakov ; Sara Rosenthal ; Pepa Atanasova ; Georgi Karadzhov ; Hamdy Mubarak ; Leon Derczynski ; Zeses Pitenis ; Çağrı Çöltekin
COMMENTS:       Proceedings of SemEval-2020
HIGHLIGHT:      We present the results and main findings of SemEval-2020 Task 12 on Multilingual Offensive Language Identification in Social Media (OffensEval 2020).

11, TITLE:      A Practical Sparse Approximation for Real Time Recurrent Learning
http://arxiv.org/abs/2006.07232
AUTHORS:        Jacob Menick ; Erich Elsen ; Utku Evci ; Simon Osindero ; Karen Simonyan ; Alex Graves
HIGHLIGHT:      We introduce the Sparse n-step Approximation (SnAp) to the RTRL influence matrix, which only keeps entries that are nonzero within n steps of the recurrent core.

12, TITLE:      Pitfalls of the Gram Loss for Neural Texture Synthesis in Light of Deep Feature Histograms
http://arxiv.org/abs/2006.07229
AUTHORS:        Eric Heitz ; Kenneth Vanhoey ; Thomas Chambon ; Laurent Belcour
COMMENTS:       10 pages, 10 figures
HIGHLIGHT:      In this paper, we propose a comprehensive study of these problems in the light of the multi-dimensional histograms of deep features.

13, TITLE:      FedGAN: Federated Generative AdversarialNetworks for Distributed Data
http://arxiv.org/abs/2006.07228
AUTHORS:        Mohammad Rasouli ; Tao Sun ; Ram Rajagopal
COMMENTS:       23 pages, 10 figures
HIGHLIGHT:      We propose Federated Generative Adversarial Network (FedGAN) for training a GAN across distributed sources of non-independent-and-identically-distributed data sources subject to communication and privacy constraints.

14, TITLE:      Local-Area-Learning Network: Meaningful Local Areas for Efficient Point Cloud Analysis
http://arxiv.org/abs/2006.07226
AUTHORS:        Qendrim Bytyqi ; Nicola Wolpert ; Elmar Schömer
HIGHLIGHT:      In this paper, we introduce the neural Local-Area-Learning Network (LocAL-Net) which places emphasis on the selection and characterization of the local areas.

15, TITLE:      Sparse and Continuous Attention Mechanisms
http://arxiv.org/abs/2006.07214
AUTHORS:        André F. T. Martins ; Marcos Treviso ; António Farinhas ; Vlad Niculae ; Mário A. T. Figueiredo ; Pedro M. Q. Aguiar
HIGHLIGHT:      Sparse and Continuous Attention Mechanisms

16, TITLE:      Branch-Cooperative OSNet for Person Re-Identification
http://arxiv.org/abs/2006.07206
AUTHORS:        Lei Zhang ; Xiaofu Wu ; Suofei Zhang ; Zirui Yin
COMMENTS:       7 pages, 3 figures and 5 tables
HIGHLIGHT:      In this paper, we propose a branch-cooperative architecture over OSNet, termed BC-OSNet, for person Re-ID.

17, TITLE:      Video Understanding as Machine Translation
http://arxiv.org/abs/2006.07203
AUTHORS:        Bruno Korbar ; Fabio Petroni ; Rohit Girdhar ; Lorenzo Torresani
HIGHLIGHT:      In this work we remove the need for negative sampling by taking a generative modeling approach that poses the objective as a translation problem between modalities.

18, TITLE:      A Face Preprocessing Approach for Improved DeepFake Detection
http://arxiv.org/abs/2006.07084
AUTHORS:        Polychronis Charitidis ; Giorgos Kordopatis-Zilos ; Symeon Papadopoulos ; Ioannis Kompatsiaris
HIGHLIGHT:      In this paper, we focus on this aspect of the DeepFake detection task and propose a preprocessing step to improve the quality of training datasets for the problem.

19, TITLE:      Optimal Allocation of Real-Time-Bidding and Direct Campaigns
http://arxiv.org/abs/2006.07070
AUTHORS:        Grégoire Jauvion ; Nicolas Grislain
HIGHLIGHT:      In this paper, we consider the problem of optimizing the revenue a web publisher gets through real-time bidding (i.e. from ads sold in real-time auctions) and direct (i.e. from ads sold through contracts agreed in advance).

20, TITLE:      Information Extraction of Clinical Trial Eligibility CriteriaYitong
http://arxiv.org/abs/2006.07296
AUTHORS:        Yitong Tseo ; M. I. Salkola ; Ahmed Mohamed ; Anuj Kumar ; Freddy Abnousi
COMMENTS:       4 pages
HIGHLIGHT:      In this paper, we investigate an information extraction (IE) approach for grounding criteria from trials in ClinicalTrials.gov to a shared knowledge base.

21, TITLE:      A Formal Language Approach to Explaining RNNs
http://arxiv.org/abs/2006.07292
AUTHORS:        Bishwamittra Ghosh ; Daniel Neider
HIGHLIGHT:      This paper presents LEXR, a framework for explaining the decision making of recurrent neural networks (RNNs) using a formal description language called Linear Temporal Logic (LTL).

22, TITLE:      Language-Conditioned Goal Generation: a New Approach to Language Grounding for RL
http://arxiv.org/abs/2006.07043
AUTHORS:        Cédric Colas ; Ahmed Akakzia ; Pierre-Yves Oudeyer ; Mohamed Chetouani ; Olivier Sigaud
HIGHLIGHT:      This paper proposes another approach: using language to condition goal generators.

23, TITLE:      Recurrent Neural Networks for Stochastic Control in Real-Time Bidding
http://arxiv.org/abs/2006.07042
AUTHORS:        Nicolas Grislain ; Nicolas Perrin ; Antoine Thabault
HIGHLIGHT:      This paper proposes an approximate solution based on a Recurrent Neural Network (RNN) architecture that is both effective and practical for implementation in a production environment.

24, TITLE:      Dutch General Public Reaction on Governmental COVID-19 Measures and Announcements in Twitter Data
http://arxiv.org/abs/2006.07283
AUTHORS:        Shihan Wang ; Marijn Schraagen ; Erik Tjong Kim Sang ; Mehdi Dastani
COMMENTS:       11 pages, 6 figures
HIGHLIGHT:      In this paper, we collect streaming data using the Twitter API starting from the COVID-19 outbreak in the Netherlands in February 2020, and track Dutch general public reactions on governmental measures and announcements.

25, TITLE:      Unmasking the Inductive Biases of Unsupervised Object Representations for Video Sequences
http://arxiv.org/abs/2006.07034
AUTHORS:        Marissa A. Weis ; Kashyap Chitta ; Yash Sharma ; Wieland Brendel ; Matthias Bethge ; Andreas Geiger ; Alexander S. Ecker
HIGHLIGHT:      In this paper, we argue that the established evaluation protocol of multi-object tracking tests precisely these perceptual qualities and we propose a new benchmark dataset based on procedurally generated video sequences.

26, TITLE:      Bandits with Partially Observable Offline Data
http://arxiv.org/abs/2006.06731
AUTHORS:        Guy Tennenholtz ; Uri Shalit ; Shie Mannor ; Yonathan Efroni
HIGHLIGHT:      We construct a linear bandit algorithm that takes advantage of the projected information, and prove regret bounds.

27, TITLE:      Is deep learning necessary for simple classification tasks?
http://arxiv.org/abs/2006.06730
AUTHORS:        Joseph D. Romano ; Trang T. Le ; Weixuan Fu ; Jason H. Moore
COMMENTS:       14 pages, 5 figures, 3 tables
HIGHLIGHT:      In spite of their successes, little guidance exists for when to choose one approach over the other in the context of specific real-world problems.

28, TITLE:      Early Detection of Retinopathy of Prematurity (ROP) in Retinal Fundus Images Via Convolutional Neural Networks
http://arxiv.org/abs/2006.06968
AUTHORS:        Xin Guo ; Yusuke Kikuchi ; Guan Wang ; Jinglin Yi ; Qiong Zou ; Rui Zhou
HIGHLIGHT:      In this study, we formulate the problem of detecting ROP in retinal fundus images in an optimization framework, and apply state-of-art convolutional neural network techniques to solve this problem.

29, TITLE:      Multi Layer Neural Networks as Replacement for Pooling Operations
http://arxiv.org/abs/2006.06969
AUTHORS:        Wolfgang Fuhl ; Enkelejda Kasneci
HIGHLIGHT:      In this work, we show that already one perceptron can be used very effectively as a pooling operation without increasing the complexity of the model.

30, TITLE:      Quantum-over-classical Advantage in Solving Multiplayer Games
http://arxiv.org/abs/2006.06965
AUTHORS:        Dmitry Kravchenko ; Kamil Khadiev ; Danil Serov ; Ruslan Kapralov
HIGHLIGHT:      We study the applicability of quantum algorithms in computational game theory and generalize some results related to Subtraction games, which are sometimes referred to as one-heap Nim games.

31, TITLE:      The eyes know it: FakeET -- An Eye-tracking Database to Understand Deepfake Perception
http://arxiv.org/abs/2006.06961
AUTHORS:        Parul Gupta ; Komal Chugh ; Abhinav Dhall ; Ramanathan Subramanian
COMMENTS:       8 pages
HIGHLIGHT:      We present \textbf{FakeET}-- an eye-tracking database to understand human visual perception of \emph{deepfake} videos.

32, TITLE:      Data Driven Prediction Architecture for Autonomous Driving and its Application on Apollo Platform
http://arxiv.org/abs/2006.06715
AUTHORS:        Kecheng Xu ; Xiangquan Xiao ; Jinghao Miao ; Qi Luo
COMMENTS:       Accepted by the 31st IEEE Intelligent Vehicles Symposium (2020)
HIGHLIGHT:      In this paper, we introduce a highly automated learning-based prediction model pipeline, which has been deployed on Baidu Apollo self-driving platform, to support different prediction learning sub-modules' data annotation, feature extraction, model training/tuning and deployment.

33, TITLE:      Understanding the Role of Training Regimes in Continual Learning
http://arxiv.org/abs/2006.06958
AUTHORS:        Seyed Iman Mirzadeh ; Mehrdad Farajtabar ; Razvan Pascanu ; Hassan Ghasemzadeh
HIGHLIGHT:      In this work, we depart from the typical approach of altering the learning algorithm to improve stability.

34, TITLE:      Decorrelated Double Q-learning
http://arxiv.org/abs/2006.06956
AUTHORS:        Gang Chen
COMMENTS:       8 pages, 5 figures
HIGHLIGHT:      Specifically, we introduce the decorrelated regularization item to reduce the correlation between value function approximators, which can lead to less biased estimation and low variance.

35, TITLE:      Does Unsupervised Architecture Representation Learning Help Neural Architecture Search?
http://arxiv.org/abs/2006.06936
AUTHORS:        Shen Yan ; Yu Zheng ; Wei Ao ; Xiao Zeng ; Mi Zhang
COMMENTS:       Technical report
HIGHLIGHT:      In this work, we find empirically that pre-training architecture representations using only neural architectures without their accuracies as labels considerably improve the downstream architecture search efficiency.

36, TITLE:      LSSL: Longitudinal Self-Supervised Learning
http://arxiv.org/abs/2006.06930
AUTHORS:        Qingyu Zhao ; Zixuan Liu ; Ehsan Adeli ; Kilian M. Pohl
HIGHLIGHT:      In this paper, we discuss the implication of repeated measures design on unsupervised learning by showing its tight conceptual connection to self-supervised learning and factor disentanglement.

37, TITLE:      Potential Field Guided Actor-Critic Reinforcement Learning
http://arxiv.org/abs/2006.06923
AUTHORS:        Weiya Ren
COMMENTS:       7 pages
HIGHLIGHT:      In this paper, we consider the problem of actor-critic reinforcement learning.

38, TITLE:      Iterate & Cluster: Iterative Semi-Supervised Action Recognition
http://arxiv.org/abs/2006.06911
AUTHORS:        Jingyuan Li ; Eli Shlizerman
COMMENTS:       for associated video, see https://www.youtube.com/watch?v=ewuoz2tt73E
HIGHLIGHT:      We propose a novel system for active semi-supervised feature-based action recognition.

39, TITLE:      Self-organization of multi-layer spiking neural networks
http://arxiv.org/abs/2006.06902
AUTHORS:        Guruprasad Raghavan ; Cong Lin ; Matt Thomson
COMMENTS:       11 pages, 4 figures
HIGHLIGHT:      To achieve this, we propose a modular tool-kit in the form of a dynamical system that can be seamlessly stacked to assemble multi-layer neural networks.

40, TITLE:      Improving GAN Training with Probability Ratio Clipping and Sample Reweighting
http://arxiv.org/abs/2006.06900
AUTHORS:        Yue Wu ; Pan Zhou ; Andrew Gordon Wilson ; Eric P. Xing ; Zhiting Hu
COMMENTS:       Code available at: https://github.com/Holmeswww/PPOGAN
HIGHLIGHT:      we propose a new variational GAN training framework which enjoys superior training stability.

41, TITLE:      Multigrid-in-Channels Architectures for Wide Convolutional Neural Networks
http://arxiv.org/abs/2006.06799
AUTHORS:        Jonathan Ephrath ; Lars Ruthotto ; Eran Treister
HIGHLIGHT:      We present a multigrid approach that combats the quadratic growth of the number of parameters with respect to the number of channels in standard convolutional neural networks (CNNs).

42, TITLE:      On Improving the Generalization of Face Recognition in the Presence of Occlusions
http://arxiv.org/abs/2006.06787
AUTHORS:        Xiang Xu ; Nikolaos Sarafianos ; Ioannis A. Kakadiaris
COMMENTS:       Technical Report
HIGHLIGHT:      In this paper, we address a key limitation of existing 2D face recognition methods: robustness to occlusions.

43, TITLE:      Tangent Space Sensitivity and Distribution of Linear Regions in ReLU Networks
http://arxiv.org/abs/2006.06780
AUTHORS:        Bálint Daróczy
COMMENTS:       14 pages, 4 figures, 2 tables
HIGHLIGHT:      In this paper we consider adversarial stability in the tangent space and suggest tangent sensitivity in order to characterize stability.

44, TITLE:      Run-time Mapping of Spiking Neural Networks to Neuromorphic Hardware
http://arxiv.org/abs/2006.06777
AUTHORS:        Adarsha Balaji ; Thibaut Marty ; Anup Das ; Francky Catthoor
COMMENTS:       Accepted in Springer Journal of Signal Processing Systems
HIGHLIGHT:      In this paper, we propose a design methodology to partition and map the neurons and synapses of online learning SNN-based applications to neuromorphic architectures at {run-time}.

45, TITLE:      One Ring to Rule Them All: Certifiably Robust Geometric Perception with Outliers
http://arxiv.org/abs/2006.06769
AUTHORS:        Heng Yang ; Luca Carlone
COMMENTS:       8 pages main results, 31 pages in total
HIGHLIGHT:      We propose a general and practical framework to design certifiable algorithms for robust geometric perception in the presence of a large amount of outliers.

46, TITLE:      Ansor : Generating High-Performance Tensor Programs for Deep Learning
http://arxiv.org/abs/2006.06762
AUTHORS:        Lianmin Zheng ; Chengfan Jia ; Minmin Sun ; Zhao Wu ; Cody Hao Yu ; Ameer Haj-Ali ; Yida Wang ; Jun Yang ; Danyang Zhuo ; Koushik Sen ; Joseph Gonzalez ; Ion Stoica
HIGHLIGHT:      We present Ansor, a tensor program generation framework for deep learning applications.

47, TITLE:      On Improving Temporal Consistency for Online Face Liveness Detection
http://arxiv.org/abs/2006.06756
AUTHORS:        Xiang Xu ; Yuanjun Xiong ; Wei Xia
COMMENTS:       technical report
HIGHLIGHT:      In this paper, we focus on improving the online face liveness detection system to enhance the security of the downstream face recognition system.

48, TITLE:      PRGFlow: Benchmarking SWAP-Aware Unified Deep Visual Inertial Odometry
http://arxiv.org/abs/2006.06753
AUTHORS:        Nitin J. Sanket ; Chahat Deep Singh ; Cornelia Fermüller ; Yiannis Aloimonos
COMMENTS:       16 pages, 13 figures, 10 tables. Under review T-RO
HIGHLIGHT:      To this end, we present a deep learning approach for visual translation estimation and loosely fuse it with an Inertial sensor for full 6DoF odometry estimation. We also present a detailed benchmark comparing different architectures, loss functions and compression methods to enable scalability.

49, TITLE:      An Unsupervised Information-Theoretic Perceptual Quality Metric
http://arxiv.org/abs/2006.06752
AUTHORS:        Sangnie Bhardwaj ; Ian Fischer ; Johannes Ballé ; Troy Chinen
COMMENTS:       Submitted to the 34th Conference on Neural Information Processing Systems (NeurIPS 2020)
HIGHLIGHT:      Tractable models of human perception have proved to be challenging to build.

50, TITLE:      Deep Convolutional Likelihood Particle Filter for Visual Tracking
http://arxiv.org/abs/2006.06746
AUTHORS:        Reza Jalil Mozhdehi ; Henry Medeiros
COMMENTS:       Accepted in Transactions on Computational Science & Computational Intelligence, 11 pages, 7 figures
HIGHLIGHT:      We propose a novel particle filter for convolutional-correlation visual trackers.

51, TITLE:      Quantum Robust Fitting
http://arxiv.org/abs/2006.06986
AUTHORS:        Tat-Jun Chin ; David Suter ; James Quach ; Shin Fang Chng
HIGHLIGHT:      In this paper, we explore the usage of quantum computers for robust fitting.

52, TITLE:      Gaze estimation problem tackled through synthetic images
http://arxiv.org/abs/2006.06740
AUTHORS:        Gonzalo Garde ; Andoni Larumbe-Bergera ; Benoît Bossavit ; Rafael Cabeza ; Sonia Porta ; Arantxa Villanueva
COMMENTS:       https://dl.acm.org/doi/abs/10.1145/3379156.3391368
HIGHLIGHT:      In this paper, we evaluate a synthetic framework to be used in the field of gaze estimation employing deep learning techniques.

53, TITLE:      Non-Negative Bregman Divergence Minimization for Deep Direct Density Ratio Estimation
http://arxiv.org/abs/2006.06979
AUTHORS:        Masahiro Kato ; Takeshi Teshima
HIGHLIGHT:      In this paper, we introduce a non-negative correction for empirical risk using only the prior knowledge of the upper bound of the density ratio.

54, TITLE:      Towards Robust Pattern Recognition: A Review
http://arxiv.org/abs/2006.06976
AUTHORS:        Xu-Yao Zhang ; Cheng-Lin Liu ; Ching Y. Suen
HIGHLIGHT:      In this paper, we present a comprehensive review of research towards robust pattern recognition from the perspective of breaking three basic and implicit assumptions: closed-world assumption, independent and identically distributed assumption, and clean and big data assumption, which form the foundation of most pattern recognition models.

55, TITLE:      Attribute analysis with synthetic dataset for person re-identification
http://arxiv.org/abs/2006.07139
AUTHORS:        Suncheng Xiang ; Yuzhuo Fu ; Guanjie You ; Ting Liu
HIGHLIGHT:      Based on it, we build a large-scale synthetic dataset, which are diversified and customized from different attributes, such as illumination and viewpoint.

56, TITLE:      Residual Force Control for Agile Human Behavior Imitation and Extended Motion Synthesis
http://arxiv.org/abs/2006.07364
AUTHORS:        Ye Yuan ; Kris Kitani
COMMENTS:       Video: https://youtu.be/XuzH1u78o1Y
HIGHLIGHT:      To overcome the dynamics mismatch, we propose a novel approach, residual force control (RFC), that augments a humanoid control policy by adding external residual forces into the action space.

57, TITLE:      Comparing Natural Language Processing Techniques for Alzheimer's Dementia Prediction in Spontaneous Speech
http://arxiv.org/abs/2006.07358
AUTHORS:        Thomas Searle ; Zina Ibrahim ; Richard Dobson
COMMENTS:       Submitted to INTERSPEECH 2020: Alzheimer's Dementia Recognition through Spontaneous Speech The ADReSS Challenge Workshop
HIGHLIGHT:      We find our top performing models to be a simple Term Frequency-Inverse Document Frequency (TF-IDF) vectoriser as input into a SVM model and a pre-trained Transformer based model `DistilBERT' when used as an embedding layer into simple linear models.

58, TITLE:      NAS-Bench-NLP: Neural Architecture Search Benchmark for Natural Language Processing
http://arxiv.org/abs/2006.07116
AUTHORS:        Nikita Klyuchnikov ; Ilya Trofimov ; Ekaterina Artemova ; Mikhail Salnikov ; Maxim Fedorov ; Evgeny Burnaev
HIGHLIGHT:      In this work, we step outside the computer vision domain by leveraging the language modeling task, which is the core of natural language processing (NLP).

59, TITLE:      Knowledge Distillation Meets Self-Supervision
http://arxiv.org/abs/2006.07114
AUTHORS:        Guodong Xu ; Ziwei Liu ; Xiaoxiao Li ; Chen Change Loy
COMMENTS:       Code is available at: https://github.com/xuguodong03/SSKD
HIGHLIGHT:      In this paper, we discuss practical ways to exploit those noisy self-supervision signals with selective transfer for distillation.

60, TITLE:      Deep Reinforcement Learning for Neural Control
http://arxiv.org/abs/2006.07352
AUTHORS:        Jimin Kim ; Eli Shlizerman
COMMENTS:       Please see the associated Video at: https://youtu.be/ixsUMfb9m_U
HIGHLIGHT:      We present a novel methodology for control of neural circuits based on deep reinforcement learning.

61, TITLE:      Robust Baggage Detection and Classification Based on Local Tri-directional Pattern
http://arxiv.org/abs/2006.07345
AUTHORS:         Shahbano ; Muhammad Abdullah ; Kashif Inayat
HIGHLIGHT:      Therefore, to overcome these shortcomings, our research proposed a detection algorithm for a human with or without carrying baggage.

62, TITLE:      GNN3DMOT: Graph Neural Network for 3D Multi-Object Tracking with Multi-Feature Learning
http://arxiv.org/abs/2006.07327
AUTHORS:        Xinshuo Weng ; Yongxin Wang ; Yunze Man ; Kris Kitani
COMMENTS:       CVPR 2020. My website for all my research works: http://www.xinshuoweng.com/
HIGHLIGHT:      In this work, we propose two techniques to improve the discriminative feature learning for MOT: (1) instead of obtaining features for each object independently, we propose a novel feature interaction mechanism by introducing the Graph Neural Network.

63, TITLE:      CPR: Classifier-Projection Regularization for Continual Learning
http://arxiv.org/abs/2006.07326
AUTHORS:        Sungmin Cha ; Hsiang Hsu ; Flavio P. Calmon ; Taesup Moon
HIGHLIGHT:      We propose a general, yet simple patch that can be applied to existing regularization-based continual learning methods called classifier-projection regularization (CPR).

64, TITLE:      Multiple-Vehicle Tracking in the Highway Using Appearance Model and Visual Object Tracking
http://arxiv.org/abs/2006.07309
AUTHORS:        Fateme Bafghi ; Bijan Shoushtarian
HIGHLIGHT:      This paper focuses on introducing an efficient novel approach with acceptable accuracy.

65, TITLE:      DECSTR: Learning Goal-Directed Abstract Behaviors using Pre-Verbal Spatial Predicates in Intrinsically Motivated Agents
http://arxiv.org/abs/2006.07185
AUTHORS:        Ahmed Akakzia ; Cédric Colas ; Pierre-Yves Oudeyer ; Mohamed Chetouani ; Olivier Sigaud
COMMENTS:       9 pages, 10 supplementary pages, 7 figures
HIGHLIGHT:      Guided by these findings from developmental psychology, we introduce a high-level state representation based on natural semantic predicates that describe spatial relations between objects and that are known to be present early in infants.

66, TITLE:      HMIC: Hierarchical Medical Image Classification, A Deep Learning Approach
http://arxiv.org/abs/2006.07187
AUTHORS:        Kamran Kowsari ; Rasoul Sali ; Lubaina Ehsan ; William Adorno ; Asad Ali ; Sean Moore ; Beatrice Amadi ; Paul Kelly ; Sana Syed ; Donald Brown
HIGHLIGHT:      This paper outlines an approach that is different from the current medical image classification tasks that view the issue as multi-class classification.

67, TITLE:      ESAD: Endoscopic Surgeon Action Detection Dataset
http://arxiv.org/abs/2006.07164
AUTHORS:        Vivek Singh Bawa ; Gurkirt Singh ; Francis KapingA ;  InnaSkarga-Bandurova ; Alice Leporini ; Carmela Landolfo ; Armando Stabile ; Francesco Setti ; Riccardo Muradore ; Elettra Oleari ; Fabio Cuzzolin
COMMENTS:       In context of SARAS ESAD Challeneg at MIDL
HIGHLIGHT:      In this work, we take aim towards increasing the effectiveness of surgical assistant robots. To this, we introduce a challenging dataset for surgeon action detection in real-world endoscopic videos.

68, TITLE:      Move-to-Data: A new Continual Learning approach with Deep CNNs, Application for image-class recognition
http://arxiv.org/abs/2006.07152
AUTHORS:        Miltiadis Poursanidis ; Jenny Benois-Pineau ; Akka Zemmari ; Boris Mansenca ; Aymar de Rugy
HIGHLIGHT:      We propose a fast continual learning layer at the end of the neuronal network.

69, TITLE:      Are we done with ImageNet?
http://arxiv.org/abs/2006.07159
AUTHORS:        Lucas Beyer ; Olivier J. Hénaff ; Alexander Kolesnikov ; Xiaohua Zhai ; Aäron van den Oord
COMMENTS:       All five authors contributed equally. New labels at https://github.com/google-research/reassessed-imagenet
HIGHLIGHT:      We therefore develop a significantly more robust procedure for collecting human annotations of the ImageNet validation set.

70, TITLE:      Algorithmic recourse under imperfect causal knowledge: a probabilistic approach
http://arxiv.org/abs/2006.06831
AUTHORS:        Amir-Hossein Karimi ; Julius von Kügelgen ; Bernhard Schölkopf ; Isabel Valera
HIGHLIGHT:      To address this limitation, we propose two probabilistic approaches to select optimal actions that achieve recourse with high probability given limited causal knowledge (e.g., only the causal graph).

71, TITLE:      Modelling Hierarchical Structure between Dialogue Policy and Natural Language Generator with Option Framework for Task-oriented Dialogue System
http://arxiv.org/abs/2006.06814
AUTHORS:        Jianhong Wang ; Yuan Zhang ; Tae-Kyun Kim ; Yunjie Gu
HIGHLIGHT:      In our work, we (1) propose modelling the hierarchical structure between dialogue policy and natural language generator (NLG) with the option framework, called HDNO; (2) train HDNO with hierarchical reinforcement learning (HRL), as well as suggest alternating updates between dialogue policy and NLG during HRL inspired by fictitious play, to preserve the comprehensibility of generated system utterances while improving fulfilling user requests; and (3) propose using a discriminator modelled with language models as an additional reward to further improve the comprehensibility.

72, TITLE:      Automated Identification of Thoracic Pathology from Chest Radiographs with Enhanced Training Pipeline
http://arxiv.org/abs/2006.06805
AUTHORS:        Adora M. DSouza ; Anas Z. Abidin ; Axel Wismüller
COMMENTS:       6 pages, 1 figure, 2 tables
HIGHLIGHT:      To this end, we investigate a deep-learning framework with novel training schemes for classification of different thoracic pathology labels from chest x-rays.

73, TITLE:      Human and Multi-Agent collaboration in a human-MARL teaming framework
http://arxiv.org/abs/2006.07301
AUTHORS:        Neda Navidi ; Francois Chabot ; Sagar Kurandwad ; Irv Lustigman ; Vincent Robert ; Gregory Szriftgiser ; Andrea Schuch
COMMENTS:       Under Review for neurips 2020
HIGHLIGHT:      This study proposes two innovative solutions to address the complexities of a collaboration between a human and multiple reinforcement learning (RL)-based agents (referred to thereafter as Human-MARL teaming) where the goals pursued cannot be achieved by a human alone or agents alone.

74, TITLE:      Recurrent Sum-Product-Max Networks for Decision Making in Perfectly-Observed Environments
http://arxiv.org/abs/2006.07300
AUTHORS:        Hari Teja Tatavarti ; Prashant Doshi ; Layton Hayes
HIGHLIGHT:      In this paper, we present recurrent SPMNs (RSPMN) that learn from and model decision-making data over time.

75, TITLE:      A New Perspective on Learning Context-Specific Independence
http://arxiv.org/abs/2006.06896
AUTHORS:        Yujia Shen ; Arthur Choi ; Adnan Darwiche
HIGHLIGHT:      In this paper, we provide a new perspective on how to learn CSIs from data.

76, TITLE:      Online Sequential Extreme Learning Machines: Features Combined From Hundreds of Midlayers
http://arxiv.org/abs/2006.06893
AUTHORS:        Chandra Swarathesh Addanki
HIGHLIGHT:      In this paper, we develop an algorithm called hierarchal online sequential learning algorithm (H-OS-ELM) for single feed feedforward network with features combined from hundreds of midlayers, the algorithm can learn chunk by chunk with fixed or varying block size, we believe that the diverse selectivity of neurons in top layers which consists of encoded distributed information produced by the other neurons offers better computational advantage over inference accuracy.

77, TITLE:      SemifreddoNets: Partially Frozen Neural Networks for Efficient Computer Vision Systems
http://arxiv.org/abs/2006.06888
AUTHORS:        Leo F Isikdogan ; Bhavin V Nayak ; Chyuan-Tyng Wu ; Joao Peralta Moreira ; Sushma Rao ; Gilad Michael
HIGHLIGHT:      We propose a system comprised of fixed-topology neural networks having partially frozen weights, named SemifreddoNets.

78, TITLE:      Rethinking Pre-training and Self-training
http://arxiv.org/abs/2006.06882
AUTHORS:        Barret Zoph ; Golnaz Ghiasi ; Tsung-Yi Lin ; Yin Cui ; Hanxiao Liu ; Ekin D. Cubuk ; Quoc V. Le
HIGHLIGHT:      Our study reveals the generality and flexibility of self-training with three additional insights: 1) stronger data augmentation and more labeled data further diminish the value of pre-training, 2) unlike pre-training, self-training is always helpful when using stronger data augmentation, in both low-data and high-data regimes, and 3) in the case that pre-training is helpful, self-training improves upon pre-training.

79, TITLE:      Reintroducing Straight-Through Estimators as Principled Methods for Stochastic Binary Networks
http://arxiv.org/abs/2006.06880
AUTHORS:        Viktor Yanush ; Alexander Shekhovtsov ; Dmitry Molchanov ; Dmitry Vetrov
HIGHLIGHT:      By choosing the activation noises in SBN appropriately and choosing mirror descent (MD) for optimization, we obtain methods that closely resemble several existing straight-through variants, but unlike them, all work reliably and produce equally good results.

80, TITLE:      High-Precision Extraction of Emerging Concepts from Scientific Literature
http://arxiv.org/abs/2006.06877
AUTHORS:        Daniel King ; Doug Downey ; Daniel S. Weld
COMMENTS:       Accepted to SIGIR 2020
HIGHLIGHT:      We present an unsupervised concept extraction method for scientific literature that achieves much higher precision than previous work. To stimulate research in this area, we release our code and data (https://github.com/allenai/ForeCite).

81, TITLE:      Learning to Play by Imitating Humans
http://arxiv.org/abs/2006.06874
AUTHORS:        Rostam Dinyari ; Pierre Sermanet ; Corey Lynch
HIGHLIGHT:      In this work, we explore the question of whether robots can learn to play to autonomously generate play data that can ultimately enhance performance.

82, TITLE:      The Smoothed Possibility of Social Choice
http://arxiv.org/abs/2006.06875
AUTHORS:        Lirong Xia
HIGHLIGHT:      We develop a framework to leverage the elegant "worst average-case" idea in smoothed complexity analysis to social choice, motivated by modern applications of social choice powered by AI and ML.

83, TITLE:      FastPitch: Parallel Text-to-speech with Pitch Prediction
http://arxiv.org/abs/2006.06873
AUTHORS:        Adrian Łańcucki
HIGHLIGHT:      We present FastPitch, a fully-parallel text-to-speech model based on FastSpeech, conditioned on fundamental frequency contours.

84, TITLE:      Multi-Agent Informational Learning Processes
http://arxiv.org/abs/2006.06870
AUTHORS:        Justin K Terry ; Nathaniel Grammel
HIGHLIGHT:      We introduce a new mathematical model of multi-agent reinforcement learning,the Multi-Agent Informational Learning Process or "MAILP" model.

85, TITLE:      Feudal Steering: Hierarchical Learning for Steering Angle Prediction
http://arxiv.org/abs/2006.06869
AUTHORS:        Faith Johnson ; Kristin Dana
COMMENTS:       InThe IEEE/CVFConference on Computer Vision and Pattern Recognition(CVPR) Workshops, June 2020
HIGHLIGHT:      In this work, we explore the use of feudal networks, used in hierarchical reinforcement learning (HRL), to devise a vehicle agent to predict steering angles from first person, dash-cam images of the Udacity driving dataset.

86, TITLE:      SegNBDT: Visual Decision Rules for Segmentation
http://arxiv.org/abs/2006.06868
AUTHORS:        Alvin Wan ; Daniel Ho ; Younjin Song ; Henk Tillman ; Sarah Adel Bargal ; Joseph E. Gonzalez
COMMENTS:       8 pages, 8 figures
HIGHLIGHT:      In this work, we build a hybrid neural-network and decision-tree model for segmentation that (1) attains neural network segmentation accuracy and (2) provides semi-automatically constructed visual decision rules such as "Is there a window?"

87, TITLE:      Few-shot Neural Architecture Search
http://arxiv.org/abs/2006.06863
AUTHORS:        Yiyang Zhao ; Linnan Wang ; Yuandong Tian ; Rodrigo Fonseca ; Tian Guo
HIGHLIGHT:      In this work, we propose few-shot NAS that explores the choice of using multiple super-nets: each super-net is pre-trained to be in charge of a sub-region of the search space.

88, TITLE:      Robustness to Adversarial Attacks in Learning-Enabled Controllers
http://arxiv.org/abs/2006.06861
AUTHORS:        Zikang Xiong ; Joe Eappen ; He Zhu ; Suresh Jagannathan
COMMENTS:       17 pages
HIGHLIGHT:      We provide a two-stage approach to construct this defense and show its effectiveness through a range of experiments on realistic continuous control domains such as the navigation control-loop of an F16 aircraft and the motion control system of humanoid robots.

89, TITLE:      Bandit-PAM: Almost Linear Time $k$-Medoids Clustering via Multi-Armed Bandits
http://arxiv.org/abs/2006.06856
AUTHORS:        Mo Tiwari ; Martin Jinye Zhang ; James Mayclin ; Sebastian Thrun ; Chris Piech ; Ilan Shomorony
COMMENTS:       18 pages
HIGHLIGHT:      We propose Bandit-PAM, a randomized algorithm inspired by techniques from multi-armed bandits, that significantly improves the computational efficiency of PAM.



==========Updates to Previous Papers==========
1, TITLE:       Learning Functionally Decomposed Hierarchies for Continuous Control Tasks with Path Planning
http://arxiv.org/abs/2002.05954
AUTHORS:        Sammy Christen ; Lukas Jendele ; Emre Aksan ; Otmar Hilliges
COMMENTS:       Preprint under review
HIGHLIGHT:      We present HiDe, a novel hierarchical reinforcement learning architecture that successfully solves long horizon control tasks and generalizes to unseen test scenarios.

2, TITLE:       Telling Left from Right: Learning Spatial Correspondence of Sight and Sound
http://arxiv.org/abs/2006.06175
AUTHORS:        Karren Yang ; Bryan Russell ; Justin Salamon
COMMENTS:       CVPR 2020
HIGHLIGHT:      We propose a novel self-supervised task to leverage an orthogonal principle: matching spatial information in the audio stream to the positions of sound sources in the visual stream. To train and evaluate our method, we introduce a large-scale video dataset, YouTube-ASMR-300K, with spatial audio comprising over 900 hours of footage.

3, TITLE:       Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization
http://arxiv.org/abs/2006.05078
AUTHORS:        Samuel Daulton ; Maximilian Balandat ; Eytan Bakshy
HIGHLIGHT:      We derive a novel formulation of $q$-Expected Hypervolume Improvement ($q$EHVI), an acquisition function that extends EHVI to the parallel, constrained evaluation setting.

4, TITLE:       LatentFusion: End-to-End Differentiable Reconstruction and Rendering for Unseen Object Pose Estimation
http://arxiv.org/abs/1912.00416
AUTHORS:        Keunhong Park ; Arsalan Mousavian ; Yu Xiang ; Dieter Fox
COMMENTS:       CVPR 2020, Project Page: https://keunhong.com/publications/latentfusion/ , Video: https://youtu.be/tlzcq1KYXd8 , Code: https://github.com/NVlabs/latentfusion . We have added experiments for LINEMOD and have updated the experiments on MOPED. We've also added more technical and implementation details to the methods section
HIGHLIGHT:      We propose a novel framework for 6D pose estimation of unseen objects.

5, TITLE:       iUNets: Fully invertible U-Nets with Learnable Up- and Downsampling
http://arxiv.org/abs/2005.05220
AUTHORS:        Christian Etmann ; Rihuan Ke ; Carola-Bibiane Schönlieb
HIGHLIGHT:      Here, we present a new fully-invertible U-Net-based architecture called the iUNet, which employs novel learnable and invertible up- and downsampling operations, thereby making the use of memory-efficient backpropagation possible.

6, TITLE:       Actor-Critic Policy Optimization in Partially Observable Multiagent Environments
http://arxiv.org/abs/1810.09026
AUTHORS:        Sriram Srinivasan ; Marc Lanctot ; Vinicius Zambaldi ; Julien Perolat ; Karl Tuyls ; Remi Munos ; Michael Bowling
COMMENTS:       NeurIPS 2018
HIGHLIGHT:      In this paper, we examine the role of these policy gradient and actor-critic algorithms in partially-observable multiagent environments.

7, TITLE:       Adversarial Attack Vulnerability of Medical Image Analysis Systems: Unexplored Factors
http://arxiv.org/abs/2006.06356
AUTHORS:        Suzanne C. Wetstein ; Cristina González-Gonzalo ; Gerda Bortsova ; Bart Liefers ; Florian Dubost ; Ioannis Katramados ; Laurens Hogeweg ; Bram van Ginneken ; Josien P. W. Pluim ; Marleen de Bruijne ; Clara I. Sánchez ; Mitko Veta
COMMENTS:       First three authors contributed equally
HIGHLIGHT:      In this paper, we study several previously unexplored factors affecting adversarial attack vulnerability of deep learning MedIA systems in three medical domains: ophthalmology, radiology and pathology.

8, TITLE:       Invariant Causal Prediction for Block MDPs
http://arxiv.org/abs/2003.06016
AUTHORS:        Amy Zhang ; Clare Lyle ; Shagun Sodhani ; Angelos Filos ; Marta Kwiatkowska ; Joelle Pineau ; Yarin Gal ; Doina Precup
COMMENTS:       Accepted to ICML 2020. 16 pages, 8 figures
HIGHLIGHT:      In this paper, we consider the problem of learning abstractions that generalize in block MDPs, families of environments with a shared latent state space and dynamics structure over that latent space, but varying observations.

9, TITLE:       Learning to Reason in Large Theories without Imitation
http://arxiv.org/abs/1905.10501
AUTHORS:        Kshitij Bansal ; Christian Szegedy ; Markus N. Rabe ; Sarah M. Loos ; Viktor Toman
COMMENTS:       Major revision
HIGHLIGHT:      In this paper, we demonstrate how to do automated theorem proving in the presence of a large knowledge base of potential premises without learning from human proofs.

10, TITLE:      Hyperspectral Image Classification with Attention Aided CNNs
http://arxiv.org/abs/2005.11977
AUTHORS:        Renlong Hang ; Zhu Li ; Qingshan Liu ; Pedram Ghamisi ; Shuvra S. Bhattacharyya
HIGHLIGHT:      Along this direction, we propose an attention aided CNN model for spectral-spatial classification of hyperspectral images.

11, TITLE:      Local Deep Implicit Functions for 3D Shape
http://arxiv.org/abs/1912.06126
AUTHORS:        Kyle Genova ; Forrester Cole ; Avneesh Sud ; Aaron Sarna ; Thomas Funkhouser
COMMENTS:       Camera ready version for CVPR 2020 Oral. Prior to review, this paper was referred to as DSIF, "Deep Structured Implicit Functions." 11 pages, 9 figures. Project video at https://youtu.be/3RAITzNWVJs
HIGHLIGHT:      The goal of this project is to learn a 3D shape representation that enables accurate surface reconstruction, compact storage, efficient computation, consistency for similar shapes, generalization across diverse shape categories, and inference from depth camera observations.

12, TITLE:      Uniform Interpolation Constrained Geodesic Learning on Data Manifold
http://arxiv.org/abs/2002.04829
AUTHORS:        Cong Geng ; Jia Wang ; Li Chen ; Wenbo Bao ; Chu Chu ; Zhiyong Gao
COMMENTS:       some experiments need to be modified
HIGHLIGHT:      In this paper, we propose a method to learn a minimizing geodesic within a data manifold.

13, TITLE:      Language as a Cognitive Tool to Imagine Goals in Curiosity-Driven Exploration
http://arxiv.org/abs/2002.09253
AUTHORS:        Cédric Colas ; Tristan Karch ; Nicolas Lair ; Jean-Michel Dussoux ; Clément Moulin-Frier ; Peter Ford Dominey ; Pierre-Yves Oudeyer
COMMENTS:       Contains main article and supplementaries
HIGHLIGHT:      We introduce Imagine, an intrinsically motivated deep reinforcement learning architecture that models this ability.

14, TITLE:      Evolving Normalization-Activation Layers
http://arxiv.org/abs/2004.02967
AUTHORS:        Hanxiao Liu ; Andrew Brock ; Karen Simonyan ; Quoc V. Le
HIGHLIGHT:      Here we propose to design them using an automated approach.

15, TITLE:      Neural Collaborative Reasoning
http://arxiv.org/abs/2005.08129
AUTHORS:        Hanxiong Chen ; Shaoyun Shi ; Yunqi Li ; Yongfeng Zhang
COMMENTS:       10 pages, 5 figures
HIGHLIGHT:      Inspired by recent progress on neural-symbolic machine learning, we propose a neural collaborative reasoning framework to integrate the power of embedding learning and logical reasoning, where the embeddings capture similarity patterns in data from perceptual perspectives, and the logic facilitates cognitive reasoning for informed decision making.

16, TITLE:      A Quantum Search Decoder for Natural Language Processing
http://arxiv.org/abs/1909.05023
AUTHORS:        Johannes Bausch ; Sathyawageeswar Subramanian ; Stephen Piddock
COMMENTS:       39 pages, 16 figures, 2 algorithms
HIGHLIGHT:      In this work, we construct a quantum algorithm to find the globally optimal parse (i.e. for infinite beam width) with high constant success probability.

17, TITLE:      Optimizing Neural Networks via Koopman Operator Theory
http://arxiv.org/abs/2006.02361
AUTHORS:        Akshunna S. Dogra ; William T Redman
COMMENTS:       13 pages, 3 figures
HIGHLIGHT:      In this work, we take the first steps in making use of this connection.

18, TITLE:      DeCLUTR: Deep Contrastive Learning for Unsupervised Textual Representations
http://arxiv.org/abs/2006.03659
AUTHORS:        John M. Giorgi ; Osvald Nitski ; Gary D. Bader ; Bo Wang
HIGHLIGHT:      We present DeCLUTR: Deep Contrastive Learning for Unsupervised Textual Representations, a self-supervised method for learning universal sentence embeddings that transfer to a wide variety of natural language processing (NLP) tasks.

19, TITLE:      Planning with Abstract Learned Models While Learning Transferable Subtasks
http://arxiv.org/abs/1912.07544
AUTHORS:        John Winder ; Stephanie Milani ; Matthew Landen ; Erebus Oh ; Shane Parr ; Shawn Squire ; Marie desJardins ; Cynthia Matuszek
COMMENTS:       Accepted at AAAI-20, 9 pages
HIGHLIGHT:      We introduce an algorithm for model-based hierarchical reinforcement learning to acquire self-contained transition and reward models suitable for probabilistic planning at multiple levels of abstraction.

20, TITLE:      Computing Approximate Equilibria in Sequential Adversarial Games by Exploitability Descent
http://arxiv.org/abs/1903.05614
AUTHORS:        Edward Lockhart ; Marc Lanctot ; Julien Pérolat ; Jean-Baptiste Lespiau ; Dustin Morrill ; Finbarr Timbers ; Karl Tuyls
COMMENTS:       IJCAI 2019, 11 pages, 1 figure
HIGHLIGHT:      In this paper, we present exploitability descent, a new algorithm to compute approximate equilibria in two-player zero-sum extensive-form games with imperfect information, by direct policy optimization against worst-case opponents.

21, TITLE:      Learning Global and Local Consistent Representations for Unsupervised Image Retrieval via Deep Graph Diffusion Networks
http://arxiv.org/abs/2001.01284
AUTHORS:        Zhiyong Dou ; Haotian Cui ; Lin Zhang ; Bo Wang
HIGHLIGHT:      In this paper, to address these limitations, we propose a novel method, Graph Diffusion Networks (GRAD-Net), that adopts graph neural networks (GNNs), a novel variant of deep learning algorithms on irregular graphs.

22, TITLE:      Fast Generation of High Fidelity RGB-D Images by Deep-Learning with Adaptive Convolution
http://arxiv.org/abs/2002.05067
AUTHORS:        Chuhua Xian ; Dongjiu Zhang ; Chengkai Dai ; Charlie C. L. Wang
HIGHLIGHT:      Using the raw data from consumer-level RGB-D cameras as input, we propose a deep-learning based approach to efficiently generate RGB-D images with completed information in high resolution.

23, TITLE:      What can robotics research learn from computer vision research?
http://arxiv.org/abs/2001.02366
AUTHORS:        Peter Corke ; Feras Dayoub ; David Hall ; John Skinner ; Niko Sünderhauf
COMMENTS:       15 pages, to appear in the proceeding of the International Symposium on Robotics Research (ISRR) 2019
HIGHLIGHT:      In this paper, we argue that the gains in computer vision are due to research methodology -- evaluation under strict constraints versus experiments; bold numbers versus videos.

24, TITLE:      AI Assisted Annotator using Reinforcement Learning
http://arxiv.org/abs/1910.02052
AUTHORS:        V. Ratna Saripalli ; Gopal Avinash ; Dibyajyoti Pati ; Michael Potter ; Charles W. Anderson
COMMENTS:       10 pages
HIGHLIGHT:      In this work, we report on the use of reinforcement learning to mimic the decision making process of annotators for medical events, to automate annotation and labelling.

25, TITLE:      A Novel Evolution Strategy with Directional Gaussian Smoothing for Blackbox Optimization
http://arxiv.org/abs/2002.03001
AUTHORS:        Jiaxin Zhang ; Hoang Tran ; Dan Lu ; Guannan Zhang
HIGHLIGHT:      We propose an improved evolution strategy (ES) using a novel nonlocal gradient operator for high-dimensional black-box optimization.

26, TITLE:      Edge Intelligence: Architectures, Challenges, and Applications
http://arxiv.org/abs/2003.12172
AUTHORS:        Dianlei Xu ; Tong Li ; Yong Li ; Xiang Su ; Sasu Tarkoma ; Tao Jiang ; Jon Crowcroft ; Pan Hui
COMMENTS:       53 pages, 37 figures, survey
HIGHLIGHT:      In this paper, we present a thorough and comprehensive survey on the literature surrounding edge intelligence.

27, TITLE:      Learning to Impute: A General Framework for Semi-supervised Learning
http://arxiv.org/abs/1912.10364
AUTHORS:        Wei-Hong Li ; Chuan-Sheng Foo ; Hakan Bilen
COMMENTS:       Semi-supervised Learning, Meta-Learning, Learning to impute
HIGHLIGHT:      In this paper, we take a more direct approach for semi-supervised learning and propose learning to impute the labels of unlabeled samples such that a network achieves better generalization when it is trained on these labels. We pose the problem in a learning-to-learn formulation which can easily be incorporated to the state-of-the-art semi-supervised techniques and boost their performance especially when the labels are limited.

28, TITLE:      Efficiently Calibrating Cable-Driven Surgical Robots with RGBD Fiducial Sensing and Recurrent Neural Networks
http://arxiv.org/abs/2003.08520
AUTHORS:        Minho Hwang ; Brijen Thananjeyan ; Samuel Paradis ; Daniel Seita ; Jeffrey Ichnowski ; Danyal Fer ; Thomas Low ; Ken Goldberg
COMMENTS:       8 pages, 11 figures, 3 tables
HIGHLIGHT:      We propose a novel approach to efficiently calibrate such robots by placing a 3D printed fiducial coordinate frames on the arm and end-effector that is tracked using RGBD sensing.

29, TITLE:      Generalized Product Quantization Network for Semi-supervised Image Retrieval
http://arxiv.org/abs/2002.11281
AUTHORS:        Young Kyun Jang ; Nam Ik Cho
COMMENTS:       10 pages, 10 figures, Computer Vision and Pattern Recognition (CVPR) 2020 accpeted paper
HIGHLIGHT:      To resolve this issue, we propose the first quantization-based semi-supervised image retrieval scheme: Generalized Product Quantization (GPQ) network.

30, TITLE:      Harmonization of diffusion MRI datasets with adaptive dictionary learning
http://arxiv.org/abs/1910.00272
AUTHORS:        Samuel St-Jean ; Max A. Viergever ; Alexander Leemans
COMMENTS:       v5 Peer review for Human Brain Mapping v4: Peer review round 2 v3: Peer reviewed version v2: Fix minor text issue + add supp materials v1: To be submitted to Neuroimage
HIGHLIGHT:      In this work, we propose a new harmonization algorithm based on adaptive dictionary learning to mitigate the unwanted variability caused by different scanner hardware while preserving the natural biological variability present in the data.

31, TITLE:      Automatic classification between COVID-19 pneumonia, non-COVID-19 pneumonia, and the healthy on chest X-ray image: combination of data augmentation methods
http://arxiv.org/abs/2006.00730
AUTHORS:        Mizuho Nishio ; Shunjiro Noguchi ; Hidetoshi Matsuo ; Takamichi Murakami
HIGHLIGHT:      Automatic classification between COVID-19 pneumonia, non-COVID-19 pneumonia, and the healthy on chest X-ray image: combination of data augmentation methods

32, TITLE:      DSU-net: Dense SegU-net for automatic head-and-neck tumor segmentation in MR images
http://arxiv.org/abs/2006.06278
AUTHORS:        Pin Tang ; Chen Zu ; Mei Hong ; Rui Yan ; Xingchen Peng ; Jianghong Xiao ; Xi Wu ; Jiliu Zhou ; Luping Zhou ; Yan Wang
COMMENTS:       This research needs to be advanced in the future
HIGHLIGHT:      In this paper, we propose a Dense SegU-net (DSU-net) framework for automatic NPC segmentation in MRI.

33, TITLE:      Formal Foundations of Continuous Graph Processing
http://arxiv.org/abs/1911.10982
AUTHORS:        Philip Dexter ; Yu David Liu ; Kenneth Chiu
HIGHLIGHT:      This paper describes CG Calculus, the first semantic foundation for continuous graph processing.

34, TITLE:      Kernel k-Groups via Hartigan's Method
http://arxiv.org/abs/1710.09859
AUTHORS:        Guilherme França ; Maria L. Rizzo ; Joshua T. Vogelstein
COMMENTS:       several improvements; connections with community detection and stochastic block model. Matches published version
HIGHLIGHT:      In this paper, we consider a formulation for the clustering problem using a weighted version of energy statistics in spaces of negative type.

35, TITLE:      Tangled up in BLEU: Reevaluating the Evaluation of Automatic Machine Translation Evaluation Metrics
http://arxiv.org/abs/2006.06264
AUTHORS:        Nitika Mathur ; Timothy Baldwin ; Trevor Cohn
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      We show that current methods for judging metrics are highly sensitive to the translations used for assessment, particularly the presence of outliers, which often leads to falsely confident conclusions about a metric's efficacy.

36, TITLE:      Counting of Teams in First-Order Team Logics
http://arxiv.org/abs/1902.00246
AUTHORS:        Anselm Haak ; Juha Kontinen ; Fabian Müller ; Heribert Vollmer ; Fan Yang
HIGHLIGHT:      In this paper we extend this study to classes beyond #P and extensions of first-order logic with team semantics.

37, TITLE:      NBDT: Neural-Backed Decision Trees
http://arxiv.org/abs/2004.00221
AUTHORS:        Alvin Wan ; Lisa Dunlap ; Daniel Ho ; Jihan Yin ; Scott Lee ; Henry Jin ; Suzanne Petryk ; Sarah Adel Bargal ; Joseph E. Gonzalez
COMMENTS:       8 pages, 7 figures
HIGHLIGHT:      Our NBDTs achieve (1) interpretability and (2) neural network accuracy: We preserve interpretable properties -- e.g., leaf purity and a non-ensembled model -- and demonstrate interpretability of model predictions both qualitatively and quantitatively.

38, TITLE:      Performance in the Courtroom: Automated Processing and Visualization of Appeal Court Decisions in France
http://arxiv.org/abs/2006.06251
AUTHORS:        Paul Boniol ; George Panagopoulos ; Christos Xypolopoulos ; Rajaa El Hamdani ; David Restrepo Amariles ; Michalis Vazirgiannis
HIGHLIGHT:      We propose metrics to rank lawyers based on their experience, wins/loss ratio and their importance in the network of lawyers.

39, TITLE:      Making Better Mistakes: Leveraging Class Hierarchies with Deep Networks
http://arxiv.org/abs/1912.09393
AUTHORS:        Luca Bertinetto ; Romain Mueller ; Konstantinos Tertikas ; Sina Samangooei ; Nicholas A. Lord
COMMENTS:       To appear at CVPR 2020. Code available at https://github.com/fiveai/making-better-mistakes
HIGHLIGHT:      In this paper, we aim to renew interest in this problem by reviewing past approaches and proposing two simple modifications of the cross-entropy loss which outperform the prior art under several metrics on two large datasets with complex class hierarchies: tieredImageNet and iNaturalist'19.

40, TITLE:      On the distance between two neural networks and the stability of learning
http://arxiv.org/abs/2002.03432
AUTHORS:        Jeremy Bernstein ; Arash Vahdat ; Yisong Yue ; Ming-Yu Liu
HIGHLIGHT:      Please find the Python code used in this paper here: https://github.com/jxbz/fromage.

41, TITLE:      Multi-hop Reading Comprehension across Documents with Path-based Graph Convolutional Network
http://arxiv.org/abs/2006.06478
AUTHORS:        Zeyun Tang ; Yongliang Shen ; Xinyin Ma ; Wei Xu ; Jiale Yu ; Weiming Lu
COMMENTS:       Accepted by IJCAI 2020 (copyright held by IJCAI)
HIGHLIGHT:      In this paper, we propose a novel approach to tackle this multi-hop reading comprehension problem.

42, TITLE:      Interpreting chest X-rays via CNNs that exploit hierarchical disease dependencies and uncertainty labels
http://arxiv.org/abs/1911.06475
AUTHORS:        Hieu H. Pham ; Tung T. Le ; Dat Q. Tran ; Dat T. Ngo ; Ha Q. Nguyen
COMMENTS:       This is a pre-print of our paper that was accepted by Neurocomputing - Its shorter version has been accepted by Medical Imaging with Deep Learning conference (MIDL 2020)
HIGHLIGHT:      This paper presents a supervised multi-label classification framework based on deep convolutional neural networks (CNNs) for predicting the risk of 14 common thoracic diseases.

43, TITLE:      Beyond User Self-Reported Likert Scale Ratings: A Comparison Model for Automatic Dialog Evaluation
http://arxiv.org/abs/2005.10716
AUTHORS:        Weixin Liang ; James Zou ; Zhou Yu
HIGHLIGHT:      To alleviate this problem, we formulate dialog evaluation as a comparison task.

44, TITLE:      Minimum Potential Energy of Point Cloud for Robust Global Registration
http://arxiv.org/abs/2006.06460
AUTHORS:        Zijie Wu ; Yaonan Wang ; Qing Zhu ; Jianxu Mao ; Haotian Wu ; Mingtao Feng ; Ajmal Mian
HIGHLIGHT:      In this paper, we propose a novel minimum gravitational potential energy (MPE)-based algorithm for global point set registration.

45, TITLE:      Learning Neural-Symbolic Descriptive Planning Models via Cube-Space Priors: The Voyage Home (to STRIPS)
http://arxiv.org/abs/2004.12850
AUTHORS:        Masataro Asai ; Christian Muise
COMMENTS:       Accepted in IJCAI 2020 main track (accept ratio 12.6%). The prequel of this paper, "The Search for STRIPS", can be found here: arXiv:1912.05492
HIGHLIGHT:      We achieved a new milestone in the difficult task of enabling agents to learn about their environment autonomously.

46, TITLE:      Autonomous Driving with Deep Learning: A Survey of State-of-Art Technologies
http://arxiv.org/abs/2006.06091
AUTHORS:        Yu Huang ; Yue Chen
HIGHLIGHT:      Due to the limited space, we focus the analysis on several key areas, i.e. 2D and 3D object detection in perception, depth estimation from cameras, multiple sensor fusion on the data, feature and task level respectively, behavior modelling and prediction of vehicle driving and pedestrian trajectories.

47, TITLE:      Multi Type Mean Field Reinforcement Learning
http://arxiv.org/abs/2002.02513
AUTHORS:        Sriram Ganapathi Subramanian ; Pascal Poupart ; Matthew E. Taylor ; Nidhi Hegde
COMMENTS:       Paper to appear in the Proceedings of International Conference on Autonomous Agents and Multi-Agent Systems (AAMAS) 2020. Revised version has some typos corrected
HIGHLIGHT:      In this paper, we extend mean field multiagent algorithms to multiple types.

48, TITLE:      Efficient Large-Scale Multi-Drone Delivery Using Transit Networks
http://arxiv.org/abs/1909.11840
AUTHORS:        Shushman Choudhury ; Kiril Solovey ; Mykel J. Kochenderfer ; Marco Pavone
COMMENTS:       Current version submitted to JAIR. Previous version appeared at IEEE ICRA 2020
HIGHLIGHT:      We present a comprehensive algorithmic framework that strives to minimize the maximum time to complete any delivery and addresses the multifaceted computational challenges of our problem through a two-layer approach.

49, TITLE:      Generating Diverse and Consistent QA pairs from Contexts with Information-Maximizing Hierarchical Conditional VAEs
http://arxiv.org/abs/2005.13837
AUTHORS:        Dong Bok Lee ; Seanie Lee ; Woo Tae Jeong ; Donghwan Kim ; Sung Ju Hwang
COMMENTS:       ACL 2020
HIGHLIGHT:      In this work, we propose a hierarchical conditional variational autoencoder (HCVAE) for generating QA pairs given unstructured texts as contexts, while maximizing the mutual information between generated QA pairs to ensure their consistency.

50, TITLE:      Defense of Word-level Adversarial Attacks via Random Substitution Encoding
http://arxiv.org/abs/2005.00446
AUTHORS:        Zhaoyang Wang ; Hongtao Wang
COMMENTS:       12 pages, 2 figures, 4 tables. Accepted as a FULL paper at KSEM 2020
HIGHLIGHT:      In this paper, we shed light on this problem and propose a novel defense framework called Random Substitution Encoding (RSE), which introduces a random substitution encoder into the training process of original neural networks.

51, TITLE:      GAN2GAN: Generative Noise Learning for Iterative Blind Denoising with Single Noisy Images
http://arxiv.org/abs/1905.10488
AUTHORS:        Sungmin Cha ; Taeeon Park ; Taesup Moon
HIGHLIGHT:      We tackle a challenging blind image denoising problem, in which only single distinct noisy images are available for training a denoiser, and no information about noise is known, except for it being zero-mean, additive, and independent of the clean image.

52, TITLE:      Cross-Channel Intragroup Sparsity Neural Network
http://arxiv.org/abs/1910.11971
AUTHORS:        Zhilin Yu ; Chao Wang ; Xin Wang ; Qing Wu ; Yong Zhao ; Xundong Wu
HIGHLIGHT:      This work introduces the cross-channel intragroup (CCI) sparsity structure, which can prevent the inference inefficiency of fine-grained pruning while maintaining outstanding model performance.

53, TITLE:      Graduated Non-Convexity for Robust Spatial Perception: From Non-Minimal Solvers to Global Outlier Rejection
http://arxiv.org/abs/1909.08605
AUTHORS:        Heng Yang ; Pasquale Antonante ; Vasileios Tzoumas ; Luca Carlone
COMMENTS:       10 pages, 5 figures, published at IEEE Robotics and Automation Letters (RA-L), 2020, Best Paper Award in Robot Vision at ICRA 2020
HIGHLIGHT:      In this paper, we enable the simultaneous use of non-minimal solvers and robust estimation by providing a general-purpose approach for robust global estimation, which can be applied to any problem where a non-minimal solver is available for the outlier-free case.

54, TITLE:      Cubical Ripser: Software for computing persistent homology of image and volume data
http://arxiv.org/abs/2005.12692
AUTHORS:        Shizuo Kaji ; Takeki Sudo ; Kazushi Ahara
HIGHLIGHT:      We introduce Cubical Ripser for computing persistent homology of image and volume data (more precisely, weighted cubical complexes).

55, TITLE:      Modular Meta-Learning with Shrinkage
http://arxiv.org/abs/1909.05557
AUTHORS:        Yutian Chen ; Abram L. Friesen ; Feryal Behbahani ; Arnaud Doucet ; David Budden ; Matthew W. Hoffman ; Nando de Freitas
COMMENTS:       33 pages (12 main, 21 supplement), under review
HIGHLIGHT:      In this work, we develop techniques based on Bayesian shrinkage to meta-learn how task-independent each module is and to regularize it accordingly.

56, TITLE:      Algorithmic Recourse: from Counterfactual Explanations to Interventions
http://arxiv.org/abs/2002.06278
AUTHORS:        Amir-Hossein Karimi ; Bernhard Schölkopf ; Isabel Valera
HIGHLIGHT:      In this work, we rely on causal reasoning to caution against the use of counterfactual explanations as a recommendable set of actions for recourse.

57, TITLE:      Automatic Discourse Segmentation: an evaluation in French
http://arxiv.org/abs/2002.04095
AUTHORS:        Rémy Saksik ; Alejandro Molina-Villegas ; Andréa Carneiro Linhares ; Juan-Manuel Torres-Moreno
COMMENTS:       7 pages, 2 figures, 2 tables
HIGHLIGHT:      In this article, we describe some discursive segmentation methods as well as a preliminary evaluation of the segmentation quality.

58, TITLE:      Variance Reduction Methods for Sublinear Reinforcement Learning
http://arxiv.org/abs/1802.09184
AUTHORS:        Sham Kakade ; Mengdi Wang ; Lin F. Yang
COMMENTS:       There is a technical issue in the analysis that is not easily fixable
HIGHLIGHT:      The main contribution is in providing a novel algorithm --- Variance-reduced Upper Confidence Q-learning (vUCQ) --- which enjoys a regret bound of $\widetilde{O}(\sqrt{HSAT} + H^5SA)$, where the $T$ is the number of time steps the agent acts in the MDP, $S$ is the number of states, $A$ is the number of actions, and $H$ is the (episodic) horizon time.

59, TITLE:      Combination of abstractive and extractive approaches for summarization of long scientific texts
http://arxiv.org/abs/2006.05354
AUTHORS:        Vladislav Tretyak ; Denis Stepanov
COMMENTS:       11 pages, 2 figures, 3 table, submitted to 23rd International Conference on Discovery Science. Fixed authors list
HIGHLIGHT:      In this research work, we present a method to generate summaries of long scientific documents that uses the advantages of both extractive and abstractive approaches.

60, TITLE:      Proving P!=NP in first-order PA
http://arxiv.org/abs/2005.10080
AUTHORS:        Rupert McCallum
HIGHLIGHT:      We show that it is provable in PA that there is an arithmetically definable sequence $\{\phi_{n}:n \in \omega\}$ of $\Pi^{0}_{2}$-sentences, such that - PRA+$\{\phi_{n}:n \in \omega\}$ is $\Pi^{0}_{2}$-sound and $\Pi^{0}_{1}$-complete - the length of $\phi_{n}$ is bounded above by a polynomial function of $n$ with positive leading coefficient - PRA+$\phi_{n+1}$ always proves 1-consistency of PRA+$\phi_{n}$.

61, TITLE:      Adaptive Graph Representation Learning for Video Person Re-identification
http://arxiv.org/abs/1909.02240
AUTHORS:        Yiming Wu ; Omar El Farouk Bourahla ; Xi Li ; Fei Wu ; Qi Tian ; Xue Zhou
COMMENTS:       10 pages, 7 figures
HIGHLIGHT:      Specifically, we exploit the pose alignment connection and the feature affinity connection to construct an adaptive structure-aware adjacency graph, which models the intrinsic relations between graph nodes.

62, TITLE:      Improving Place Recognition Using Dynamic Object Detection
http://arxiv.org/abs/2002.04698
AUTHORS:        Juan Pablo Munoz ; Scott Dexter
HIGHLIGHT:      We present a novel approach to place recognition well-suited to environments with many dynamic objects--objects that may or may not be present in an agent's subsequent visits.

63, TITLE:      Combining imagination and heuristics to learn strategies that generalize
http://arxiv.org/abs/1809.03406
AUTHORS:        Erik J Peterson ; Necati Alp Müyesser ; Timothy Verstynen ; Kyle Dunovan
HIGHLIGHT:      Motivated by theories of the hierarchical organization of the human prefrontal networks, we have developed a model of hierarchical reinforcement learning that combines both heuristics and imagination into a stumbler-strategist network.

64, TITLE:      Rotation, Translation, and Cropping for Zero-Shot Generalization
http://arxiv.org/abs/2001.09908
AUTHORS:        Chang Ye ; Ahmed Khalifa ; Philip Bontrager ; Julian Togelius
COMMENTS:       IEEE Conference on Games 2020 Full Paper
HIGHLIGHT:      This paper advances the hypothesis that the lack of generalization is partly due to the input representation, and explores how rotation, cropping and translation could increase generality.

65, TITLE:      Orthogonal Over-Parameterized Training
http://arxiv.org/abs/2004.04690
AUTHORS:        Weiyang Liu ; Rongmei Lin ; Zhen Liu ; James M. Rehg ; Li Xiong ; Adrian Weller ; Le Song
COMMENTS:       Technical Report v2 (35 pages)
HIGHLIGHT:      We propose a novel orthogonal over-parameterized training (OPT) framework that can provably minimize the hyperspherical energy which characterizes the diversity of neurons on a hypersphere.

66, TITLE:      Parameterizing Branch-and-Bound Search Trees to Learn Branching Policies
http://arxiv.org/abs/2002.05120
AUTHORS:        Giulia Zarpellon ; Jason Jo ; Andrea Lodi ; Yoshua Bengio
HIGHLIGHT:      We aim instead at learning a policy that generalizes across heterogeneous MILPs: our main hypothesis is that parameterizing the state of the B&B search tree can aid this type of generalization.

67, TITLE:      Neural Random Subspace
http://arxiv.org/abs/1911.07845
AUTHORS:        Yun-Hao Cao ; Jianxin Wu ; Hanchen Wang ; Joan Lasenby
COMMENTS:       34 pages
HIGHLIGHT:      We propose Neural Random Subspace (NRS), a novel deep learning based random subspace method.

68, TITLE:      Transformers Generalize to the Semantics of Logics
http://arxiv.org/abs/2003.04218
AUTHORS:        Christopher Hahn ; Frederik Schmitt ; Jens U. Kreber ; Markus N. Rabe ; Bernd Finkbeiner
HIGHLIGHT:      We show that neural networks can learn the semantics of propositional and linear-time temporal logic (LTL) from imperfect training data.

69, TITLE:      NASA: Neural Articulated Shape Approximation
http://arxiv.org/abs/1912.03207
AUTHORS:        Boyang Deng ; JP Lewis ; Timothy Jeruzalski ; Gerard Pons-Moll ; Geoffrey Hinton ; Mohammad Norouzi ; Andrea Tagliasacchi
HIGHLIGHT:      This paper introduces neural articulated shape approximation (NASA), an alternative framework that enables efficient representation of articulated deformable objects using neural indicator functions that are conditioned on pose.

70, TITLE:      PFCNN: Convolutional Neural Networks on 3D Surfaces Using Parallel Frames
http://arxiv.org/abs/1808.04952
AUTHORS:        Yuqi Yang ; Shilin Liu ; Hao Pan ; Yang Liu ; Xin Tong
COMMENTS:       15 pages, 18 figures. CVPR 2020. Project page: https://haopan.github.io/surfacecnn.html
HIGHLIGHT:      We use parallel frames on surface to define PFCNNs that enable effective feature learning on surface meshes by mimicking standard convolutions faithfully.

71, TITLE:      XAI for Graphs: Explaining Graph Neural Network Predictions by Identifying Relevant Walks
http://arxiv.org/abs/2006.03589
AUTHORS:        Thomas Schnake ; Oliver Eberle ; Jonas Lederer ; Shinichi Nakajima ; Kristof T. Schütt ; Klaus-Robert Müller ; Grégoire Montavon
COMMENTS:       12 pages + 12 pages supplement
HIGHLIGHT:      In this paper, we contribute by proposing a new XAI approach for GNNs.