2020.06.17.txt

==========New Papers==========
1, TITLE:       Visual Identification of Individual Holstein Friesian Cattle via Deep Metric Learning
http://arxiv.org/abs/2006.09205
AUTHORS:        William Andrew ; Jing Gao ; Neill Campbell ; Andrew W Dowsey ; Tilo Burghardt
COMMENTS:       35 pages, 14 figures, 2 tables
HIGHLIGHT:      Using agriculturally relevant top-down imaging, we present methods for the detection, localisation, and identification of individual Holstein Friesians in an open herd setting, i.e. where changes in the herd do not require system re-training.

2, TITLE:       On the use of human reference data for evaluating automatic image descriptions
http://arxiv.org/abs/2006.08792
AUTHORS:        Emiel van Miltenburg
COMMENTS:       Originally presented as a (non-archival) poster at the VizWiz 2020 workshop, collocated with CVPR 2020. See: https://vizwiz.org/workshops/2020-workshop/
HIGHLIGHT:      I argue that there is a need for more detailed guidelines that take into account the needs of visually impaired users, but also the feasibility of generating suitable descriptions.

3, TITLE:       Equilibrium Propagation for Complete Directed Neural Networks
http://arxiv.org/abs/2006.08798
AUTHORS:        Matilde Tristany ; Sérgio Pequito ; Pedro A. Santos ; Mário A. T. Figueiredo
COMMENTS:       6 pages, 6 images, accepted for ESANN 2020
HIGHLIGHT:      Specifically, we introduce: a new neuronal dynamics and learning rule for arbitrary network architectures; a sparsity-inducing method able to prune irrelevant connections; a dynamical-systems characterization of the models, using Lyapunov theory.

4, TITLE:       Total Deep Variation: A Stable Regularizer for Inverse Problems
http://arxiv.org/abs/2006.08789
AUTHORS:        Erich Kobler ; Alexander Effland ; Karl Kunisch ; Thomas Pock
COMMENTS:       30 pages, 12 figures. arXiv admin note: text overlap with arXiv:2001.05005
HIGHLIGHT:      In this work, we combine the variational formulation of inverse problems with deep learning by introducing the data-driven general-purpose total deep variation regularizer.

5, TITLE:       On Effective Parallelization of Monte Carlo Tree Search
http://arxiv.org/abs/2006.08785
AUTHORS:        Anji Liu ; Yitao Liang ; Ji Liu ; Guy Van den Broeck ; Jianshu Chen
HIGHLIGHT:      To this end, we propose a general parallel MCTS framework that can be specialized to major existing parallel MCTS algorithms.

6, TITLE:       Automatic Validation of Textual Attribute Values in E-commerce Catalog by Learning with Limited Labeled Data
http://arxiv.org/abs/2006.08779
AUTHORS:        Yaqing Wang ; Yifan Ethan Xu ; Xian Li ; Xin Luna Dong ; Jing Gao
COMMENTS:       Proceedings of the 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, August 23--27, 2020, Virtual Event, CA, USA
HIGHLIGHT:      To address the aforementioned challenges, we propose a novel meta-learning latent variable approach, called MetaBridge, which can learn transferable knowledge from a subset of categories with limited labeled data and capture the uncertainty of never-seen categories with unlabeled data.

7, TITLE:       COMPOSE: Cross-Modal Pseudo-Siamese Network for Patient Trial Matching
http://arxiv.org/abs/2006.08765
AUTHORS:        Junyi Gao ; Cao Xiao ; Lucas M. Glass ; Jimeng Sun
COMMENTS:       Accepted by KDD'20
HIGHLIGHT:      In this paper, we proposed CrOss-Modal PseudO-SiamEse network (COMPOSE) to address these challenges for patient-trial matching.

8, TITLE:       Towards Understanding the Effect of Leak in Spiking Neural Networks
http://arxiv.org/abs/2006.08761
AUTHORS:        Sayeed Shafayet Chowdhury ; Chankyu Lee ; Kaushik Roy
COMMENTS:       Sayeed Shafayet Chowdhury and Chankyu Lee contributed equally
HIGHLIGHT:      In this paper, we investigate the questions regarding the justification of leak and the pros and cons of using leaky behavior.

9, TITLE:       Improved Deep Point Cloud Geometry Compression
http://arxiv.org/abs/2006.09043
AUTHORS:        Maurice Quach ; Giuseppe Valenzise ; Frederic Dufaux
HIGHLIGHT:      In this paper, we propose a set of contributions to improve deep point cloud compression, i.e.: using a scale hyperprior model for entropy coding; employing deeper transforms; a different balancing weight in the focal loss; optimal thresholding for decoding; and sequential model training.

10, TITLE:      Fine-Tuning DARTS for Image Classification
http://arxiv.org/abs/2006.09042
AUTHORS:        Muhammad Suhaib Tanveer ; Muhammad Umar Karim Khan ; Chong-Min Kyung
COMMENTS:       8 pages, 6 figures
HIGHLIGHT:      We propose to fine-tune DARTS using fixed operations as they are independent of these approximations.

11, TITLE:      Debona: Decoupled Boundary Network Analysis for Tighter Bounds and Faster Adversarial Robustness Proofs
http://arxiv.org/abs/2006.09040
AUTHORS:        Christopher Brix ; Thomas Noll
HIGHLIGHT:      We propose an improved technique for computing tight upper and lower bounds of these node values, based on increased flexibility gained by computing both bounds independently of each other.

12, TITLE:      Multi-Precision Policy Enforced Training (MuPPET): A precision-switching strategy for quantised fixed-point training of CNNs
http://arxiv.org/abs/2006.09049
AUTHORS:        Aditya Rajagopal ; Diederik Adriaan Vink ; Stylianos I. Venieris ; Christos-Savvas Bouganis
COMMENTS:       Accepted at the 37th International Conference on Machine Learning (ICML), 2020
HIGHLIGHT:      This work pushes the boundary of quantised training by employing a multilevel optimisation approach that utilises multiple precisions including low-precision fixed-point representations.

13, TITLE:      On the Computational Power of Transformers and Its Implications in Sequence Modeling
http://arxiv.org/abs/2006.09286
AUTHORS:        Satwik Bhattamishra ; Arkil Patel ; Navin Goyal
HIGHLIGHT:      In this paper, we take a step towards answering these questions.

14, TITLE:      Communicative need modulates competition in language change
http://arxiv.org/abs/2006.09277
AUTHORS:        Andres Karjus ; Richard A. Blythe ; Simon Kirby ; Kenny Smith
HIGHLIGHT:      We introduce a general method for quantifying competition between linguistic elements in diachronic corpora which does not require language-specific resources other than a sufficiently large corpus.

15, TITLE:      Deep Learning based Segmentation of Fish in Noisy Forward Looking MBES Images
http://arxiv.org/abs/2006.09034
AUTHORS:        Jesper Haahr Christensen ; Lars Valdemar Mogensen ; Ole Ravn
HIGHLIGHT:      In this work, we investigate a Deep Learning (DL) approach to fish segmentation in a small dataset of noisy low-resolution images generated by a forward-looking multibeam echosounder (MBES).

16, TITLE:      The SPPD System for Schema Guided Dialogue State Tracking Challenge
http://arxiv.org/abs/2006.09035
AUTHORS:        Miao Li ; Haoqi Xiong ; Yunbo Cao
HIGHLIGHT:      This paper introduces one of our group's work on the Dialog System Technology Challenges 8 (DSTC8), the SPPD system for Schema Guided dialogue state tracking challenge.

17, TITLE:      How Secure is Distributed Convolutional Neural Network on IoT Edge Devices?
http://arxiv.org/abs/2006.09276
AUTHORS:        Hawzhin Mohammed ; Tolulope A. Odetola ; Syed Rafay Hasan
HIGHLIGHT:      In this paper, we propose Trojan attacks on CNN deployed across a distributed edge network across different nodes.

18, TITLE:      Real-time Universal Style Transfer on High-resolution Images via Zero-channel Pruning
http://arxiv.org/abs/2006.09029
AUTHORS:        Jie An ; Tao Li ; Haozhi Huang ; Li Shen ; Xuan Wang ; Yongyi Tang ; Jinwen Ma ; Wei Liu ; Jiebo Luo
HIGHLIGHT:      In this work, we propose a lightweight alternative architecture - ArtNet, which is based on GoogLeNet, and later pruned by a novel channel pruning method named Zero-channel Pruning specially designed for style transfer approaches.

19, TITLE:      Structured and Localized Image Restoration
http://arxiv.org/abs/2006.09261
AUTHORS:        Thomas Eboli ; Alex Nowak-Vila ; Jian Sun ; Francis Bach ; Jean Ponce ; Alessandro Rudi
HIGHLIGHT:      We present a novel approach to image restoration that leverages ideas from localized structured prediction and non-linear multi-task learning.

20, TITLE:      Improved Techniques for Training Score-Based Generative Models
http://arxiv.org/abs/2006.09011
AUTHORS:        Yang Song ; Stefano Ermon
HIGHLIGHT:      We provide a new theoretical analysis of learning and sampling from score models in high dimensional spaces, explaining existing failure modes and motivating new solutions that generalize across datasets.

21, TITLE:      Lio -- A Personal Robot Assistant for Human-Robot Interaction and Care Applications
http://arxiv.org/abs/2006.09019
AUTHORS:        Justinas Miseikis ; Pietro Caroni ; Patricia Duchamp ; Alina Gasser ; Rastislav Marko ; Nelija Miseikiene ; Frederik Zwilling ; Charles de Castelbajac ; Lucas Eicher ; Michael Fruh ; Hansruedi Fruh
COMMENTS:       Accepted submission at IEEE Robotics and Automation Letters (RA-L), submitted to IEEE IROS 2020
HIGHLIGHT:      Lio -- A Personal Robot Assistant for Human-Robot Interaction and Care Applications

22, TITLE:      RL-CycleGAN: Reinforcement Learning Aware Simulation-To-Real
http://arxiv.org/abs/2006.09001
AUTHORS:        Kanishka Rao ; Chris Harris ; Alex Irpan ; Sergey Levine ; Julian Ibarz ; Mohi Khansari
COMMENTS:       Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2020)
HIGHLIGHT:      In this paper, we introduce the RL-scene consistency loss for image translation, which ensures that the translation operation is invariant with respect to the Q-values associated with the image.

23, TITLE:      How Much Can I Trust You? -- Quantifying Uncertainties in Explaining Neural Networks
http://arxiv.org/abs/2006.09000
AUTHORS:        Kirill Bykov ; Marina M. -C. Höhne ; Klaus-Robert Müller ; Shinichi Nakajima ; Marius Kloft
COMMENTS:       12 pages, 10 figures
HIGHLIGHT:      We therefore contribute by proposing a new framework that allows to convert any arbitrary explanation method for neural networks into an explanation method for Bayesian neural networks, with an in-built modeling of uncertainties.

24, TITLE:      Two-Dimensional Non-Line-of-Sight Scene Estimation from a Single Edge Occluder
http://arxiv.org/abs/2006.09241
AUTHORS:        Sheila W. Seidel ; John Murray-Bruce ; Yanting Ma ; Christopher Yu ; William T. Freeman ; Vivek K Goyal
COMMENTS:       14 pages, 15 figures
HIGHLIGHT:      We derive a new forward model, accounting for radial falloff, and propose two inversion algorithms to form 2D reconstructions from a single photograph of the penumbra.

25, TITLE:      Modeling Graph Structure via Relative Position for Better Text Generation from Knowledge Graphs
http://arxiv.org/abs/2006.09242
AUTHORS:        Martin Schmitt ; Leonardo F. R. Ribeiro ; Philipp Dufter ; Iryna Gurevych ; Hinrich Schütze
HIGHLIGHT:      We present a novel encoder-decoder architecture for graph-to-text generation based on Transformer, called the Graformer.

26, TITLE:      AcED: Accurate and Edge-consistent Monocular Depth Estimation
http://arxiv.org/abs/2006.09243
AUTHORS:        Kunal Swami ; Prasanna Vishnu Bondada ; Pankaj Kumar Bajpai
COMMENTS:       Accepted in IEEE ICIP 2020
HIGHLIGHT:      The current state-of-the-art method formulates the problem as that of ordinal regression.

27, TITLE:      Model Embedding Model-Based Reinforcement Learning
http://arxiv.org/abs/2006.09234
AUTHORS:        Xiaoyu Tan ; Chao Qu ; Junwu Xiong ; James Zhang
HIGHLIGHT:      In this paper, we propose a simple and elegant model-embedding model-based reinforcement learning (MEMB) algorithm in the framework of the probabilistic reinforcement learning.

28, TITLE:      Weakly-supervised Domain Adaption for Aspect Extraction via Multi-level Interaction Transfer
http://arxiv.org/abs/2006.09235
AUTHORS:        Tao Liang ; Wenya Wang ; Fengmao Lv
COMMENTS:       This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
HIGHLIGHT:      To this end, we propose a novel multi-level reconstruction mechanism that aligns both the fine-grained and coarse-grained information in multiple levels of abstractions.

29, TITLE:      Foreground-Background Imbalance Problem in Deep Object Detectors: A Review
http://arxiv.org/abs/2006.09238
AUTHORS:        Joya Chen ; Qi Wu ; Dong Liu ; Tong Xu
COMMENTS:       Accepted by IEEE MIPR 2020
HIGHLIGHT:      In this paper, we survey the recent advances about the solutions to the imbalance problem.

30, TITLE:      Towards Automated Assessment of Stuttering and Stuttering Therapy
http://arxiv.org/abs/2006.09222
AUTHORS:        Sebastian P. Bayerl ; Florian Hönig ; Joelle Reister ; Korbinian Riedhammer
COMMENTS:       10 pages, 3 figures, 1 table Accepted at TSD 2020, 23rd International Conference on Text, Speech and Dialogue
HIGHLIGHT:      This paper introduces the Speech Control Index (SCI), a new method to evaluate the severity of stuttering.

31, TITLE:      DSDANet: Deep Siamese Domain Adaptation Convolutional Neural Network for Cross-domain Change Detection
http://arxiv.org/abs/2006.09225
AUTHORS:        Hongruixuan Chen ; Chen Wu ; Bo Du ; Liangpei Zhang
HIGHLIGHT:      In this paper, we propose a novel deep siamese domain adaptation convolutional neural network (DSDANet) architecture for cross-domain CD.

32, TITLE:      Parameter-based Value Functions
http://arxiv.org/abs/2006.09226
AUTHORS:        Francesco Faccio ; Jürgen Schmidhuber
COMMENTS:       24 pages, 67 references
HIGHLIGHT:      We introduce a class of value functions called Parameter-based Value Functions (PVFs) whose inputs include the policy parameters.

33, TITLE:      MS-TCN++: Multi-Stage Temporal Convolutional Network for Action Segmentation
http://arxiv.org/abs/2006.09220
AUTHORS:        Shijie Li ; Yazan Abu Farha ; Yun Liu ; Ming-Ming Cheng ; Juergen Gall
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:1903.01945
HIGHLIGHT:      In this paper, we propose a multi-stage architecture for the temporal action segmentation task that overcomes the limitations of the previous approaches.

34, TITLE:      Preference-based Reinforcement Learning with Finite-Time Guarantees
http://arxiv.org/abs/2006.08910
AUTHORS:        Yichong Xu ; Ruosong Wang ; Lin F. Yang ; Aarti Singh ; Artur Dubrawski
COMMENTS:       22 pages, 2 figures
HIGHLIGHT:      If preferences are stochastic, and the preference probability relates to the hidden reward values, we present algorithms for PbRL, both with and without a simulator, that are able to identify the best policy up to accuracy $\varepsilon$ with high probability.

35, TITLE:      META-Learning Eligibility Traces for More Sample Efficient Temporal Difference Learning
http://arxiv.org/abs/2006.08906
AUTHORS:        Mingde Zhao
COMMENTS:       A thesis submitted to McGill University in partial fulfillment of the requirements of the degree of Master of Computer Science
HIGHLIGHT:      To improve the sample efficiency of TD-learning, we propose a meta-learning method for adjusting the eligibility trace parameter, in a state-dependent manner.

36, TITLE:      Causal Knowledge Extraction from Scholarly Papers in Social Sciences
http://arxiv.org/abs/2006.08904
AUTHORS:        Victor Zitian Chen ; Felipe Montano-Campos ; Wlodek Zadrozny
HIGHLIGHT:      In this paper, we seek to develop natural language processing (NLP) models to accelerate the speed of extraction of relationships from scholarly papers in social sciences, identify hypotheses from these papers, and extract the cause-and-effect entities.

37, TITLE:      Depth by Poking: Learning to Estimate Depth from Self-Supervised Grasping
http://arxiv.org/abs/2006.08903
AUTHORS:        Ben Goodrich ; Alex Kuefler ; William D. Richards
COMMENTS:       IEEE International Conference on Robotics and Automation (ICRA) 2020
HIGHLIGHT:      We address this problem by training a neural network model to estimate depth from RGB-D images, using labels from physical interactions between a robot and its environment.

38, TITLE:      AVLnet: Learning Audio-Visual Language Representations from Instructional Videos
http://arxiv.org/abs/2006.09199
AUTHORS:        Andrew Rouditchenko ; Angie Boggust ; David Harwath ; Dhiraj Joshi ; Samuel Thomas ; Kartik Audhkhasi ; Rogerio Feris ; Brian Kingsbury ; Michael Picheny ; Antonio Torralba ; James Glass
HIGHLIGHT:      In this work, we introduce Audio-Video Language Network (AVLnet), a self-supervised network that learns a shared audio-visual embedding space directly from raw video inputs.

39, TITLE:      Reconstruction of turbulent data with deep generative models for semantic inpainting from TURB-Rot database
http://arxiv.org/abs/2006.09179
AUTHORS:        M. Buzzicotti ; F. Bonaccorso ; P. Clark Di Leoni ; L. Biferale
HIGHLIGHT:      We present two approaches both based on Context Encoders.

40, TITLE:      Results of the seventh edition of the BioASQ Challenge
http://arxiv.org/abs/2006.09174
AUTHORS:        Anastasios Nentidis ; Konstantinos Bougiatiotis ; Anastasia Krithara ; Georgios Paliouras
COMMENTS:       17 pages, 2 figures
HIGHLIGHT:      The results of the seventh edition of the BioASQ challenge are presented in this paper.

41, TITLE:      Pessimism About Unknown Unknowns Inspires Conservatism
http://arxiv.org/abs/2006.08753
AUTHORS:        Michael K. Cohen ; Marcus Hutter
COMMENTS:       12 pages, plus 16-page appendix; to be published in COLT 2020 proceedings
HIGHLIGHT:      In high-stakes environments, we might like advanced artificial agents to pursue goals cautiously, which is a non-trivial problem even if the agent were allowed arbitrary computing power; we present a formal solution.

42, TITLE:      DynE: Dynamic Ensemble Decoding for Multi-Document Summarization
http://arxiv.org/abs/2006.08748
AUTHORS:        Chris Hokamp ; Demian Gholipour Ghalandari ; Nghia The Pham ; John Glover
HIGHLIGHT:      In this work we propose a simple decoding methodology which ensembles the output of multiple instances of the same model on different inputs.

43, TITLE:      A Critical and Moving-Forward View on Quantum Image Processing
http://arxiv.org/abs/2006.08747
AUTHORS:        Fei Yan ; Salvador E. Venegas-Andraca ; Kaoru Hirota
COMMENTS:       16 pages, 4 figures. Under review
HIGHLIGHT:      A Critical and Moving-Forward View on Quantum Image Processing

44, TITLE:      Certifying Strategyproof Auction Networks
http://arxiv.org/abs/2006.08742
AUTHORS:        Michael J. Curry ; Ping-Yeh Chiang ; Tom Goldstein ; John Dickerson
HIGHLIGHT:      We propose ways to explicitly verify strategyproofness under a particular valuation profile using techniques from the neural network verification literature.

45, TITLE:      Exact and Metaheuristic Approaches for the Production Leveling Problem
http://arxiv.org/abs/2006.08731
AUTHORS:        Johannes Vass ; Marie-Louise Lackner ; Nysret Musliu
COMMENTS:       Instance set is published under https://dbai.tuwien.ac.at/staff/jvass/production-leveling/
HIGHLIGHT:      In this paper we introduce a new problem in the field of production planning which we call the Production Leveling Problem.

46, TITLE:      On the Hardness of Problems Involving Negator Relationships in an Artificial Hormone System
http://arxiv.org/abs/2006.08958
AUTHORS:        Eric Hutter ; Mathias Pacher ; Uwe Brinkschulte
HIGHLIGHT:      In this supplementary report to these papers, we show examples of Negator-Path and Negator-Sat, introduce the novel problem Negator-Stability and explain why all of these problems involving negators are hard to solve algorithmically.

47, TITLE:      Latent Bandits Revisited
http://arxiv.org/abs/2006.08714
AUTHORS:        Joey Hong ; Branislav Kveton ; Manzil Zaheer ; Yinlam Chow ; Amr Ahmed ; Craig Boutilier
COMMENTS:       16 pages, 2 figures
HIGHLIGHT:      In this work, we propose general algorithms for this setting, based on both upper confidence bounds (UCBs) and Thompson sampling.

48, TITLE:      Manipulating emotions for ground truth emotion analysis
http://arxiv.org/abs/2006.08952
AUTHORS:        Bennett Kleinberg
COMMENTS:       preprint
HIGHLIGHT:      As an alternative, this paper introduces online emotion induction techniques from experimental behavioural research as a method for text-based emotion analysis.

49, TITLE:      HyperFlow: Representing 3D Objects as Surfaces
http://arxiv.org/abs/2006.08710
AUTHORS:        Przemysław Spurek ; Maciej Zięba ; Jacek Tabor ; Tomasz Trzciński
HIGHLIGHT:      In this work, we present HyperFlow - a novel generative model that leverages hypernetworks to create continuous 3D object representations in a form of lightweight surfaces (meshes), directly out of point clouds.

50, TITLE:      SPLASH: Learnable Activation Functions for Improving Accuracy and Adversarial Robustness
http://arxiv.org/abs/2006.08947
AUTHORS:        Mohammadamin Tavakoli ; Forest Agostinelli ; Pierre Baldi
HIGHLIGHT:      We introduce SPLASH units, a class of learnable activation functions shown to simultaneously improve the accuracy of deep neural networks while also improving their robustness to adversarial attacks.

51, TITLE:      Global Feature Aggregation for Accident Anticipation
http://arxiv.org/abs/2006.08942
AUTHORS:        Mishal Fatima ; Muhammad Umar Karim Khan ; Chong Min Kyung
HIGHLIGHT:      We propose a novel Feature Aggregation (FA) block that refines each object's features by computing a weighted sum of the features of all objects in a frame.

52, TITLE:      Learning the Redundancy-free Features for Generalized Zero-Shot Object Recognition
http://arxiv.org/abs/2006.08939
AUTHORS:        Zongyan Han ; Zhenyong Fu ; Jian Yang
COMMENTS:       Accepted to IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020
HIGHLIGHT:      To reduce the superfluous information in the fine-grained objects, in this paper, we propose to learn the redundancy-free features for generalized zero-shot learning.

53, TITLE:      Channel Relationship Prediction with Forget-Update Module for Few-shot Classification
http://arxiv.org/abs/2006.08937
AUTHORS:        Minglei Yuan ; Cunhao Cai ; Tong Lu
HIGHLIGHT:      In this paper, we proposed a pipeline for inferring the relationship of each class in support set and a query sample using forget-update module.

54, TITLE:      Plug-and-Play Anomaly Detection with Expectation Maximization Filtering
http://arxiv.org/abs/2006.08933
AUTHORS:        Muhammad Umar Karim Khan ; Mishal Fatima ; Chong-Min Kyung
HIGHLIGHT:      We tackle all these constraints with our approach in this paper.

55, TITLE:      Computing Igusa's local zeta function of univariates in deterministic polynomial-time
http://arxiv.org/abs/2006.08926
AUTHORS:        Ashish Dwivedi ; Nitin Saxena
COMMENTS:       15 pages, ANTS 2020
HIGHLIGHT:      We give an elementary proof of this fact for a univariate polynomial $f$.

56, TITLE:      GCNs-Net: A Graph Convolutional Neural Network Approach for Decoding Time-resolved EEG Motor Imagery Signals
http://arxiv.org/abs/2006.08924
AUTHORS:        Xiangmin Lun ; Shuyue Jia ; Yimin Hou ; Yan Shi ; Yang Li ; Hanrui Yang ; Shu Zhang ; Jinglei Lv
HIGHLIGHT:      To fill the gap, a novel deep learning framework based on the graph convolutional neural networks (GCNs) was presented to enhance the decoding performance of raw EEG signals during different types of motor imagery (MI) tasks while cooperating with the functional topological relationship of electrodes.

57, TITLE:      Domain Adaptation with Morphologic Segmentation
http://arxiv.org/abs/2006.09322
AUTHORS:        Jonathan Klein ; Sören Pirk ; Dominik L. Michels
COMMENTS:       This work has been supported by KAUST under individual baseline funding
HIGHLIGHT:      We present a novel domain adaptation framework that uses morphologic segmentation to translate images from arbitrary input domains (real and synthetic) into a uniform output domain.

58, TITLE:      The Teaching Dimension of Q-learning
http://arxiv.org/abs/2006.09324
AUTHORS:        Xuezhou Zhang ; Shubham Kumar Bharti ; Yuzhe Ma ; Adish Singla ; Xiaojin Zhu
HIGHLIGHT:      In this paper, we initiate the study of sample complexity of teaching, termed as "teaching dimension" (TDim) in the literature, for Q-learning.

59, TITLE:      Deep Multimodal Transfer-Learned Regression in Data-Poor Domains
http://arxiv.org/abs/2006.09310
AUTHORS:        Levi McClenny ; Mulugeta Haile ; Vahid Attari ; Brian Sadler ; Ulisses Braga-Neto ; Raymundo Arroyave
HIGHLIGHT:      Here we propose a Deep Multimodal Transfer-Learned Regressor (DMTL-R) for multimodal learning of image and feature data in a deep regression architecture effective at predicting target parameters in data-poor domains.

60, TITLE:      Lung Segmentation and Nodule Detection in Computed Tomography Scan using a Convolutional Neural Network Trained Adversarially using Turing Test Loss
http://arxiv.org/abs/2006.09308
AUTHORS:        Rakshith Sathish ; Rachana Sathish ; Ramanathan Sethuraman ; Debdoot Sheet
COMMENTS:       Accepted at 42nd Annual International Conferences of the IEEE Engineering in Medicine and Biology Society (2020)
HIGHLIGHT:      To tackle this problem we propose a computationally efficient two stage framework.

61, TITLE:      Unsupervised Pansharpening Based on Self-Attention Mechanism
http://arxiv.org/abs/2006.09303
AUTHORS:        Ying Qu ; Razieh Kaviani Baghbaderani ; Hairong Qi ; Chiman Kwan
COMMENTS:       submitted to TGRS
HIGHLIGHT:      In this paper, we propose an unsupervised pansharpening (UP) method in a deep-learning framework to address the above challenges based on the self-attention mechanism (SAM), referred to as UP-SAM.

62, TITLE:      Learning About Objects by Learning to Interact with Them
http://arxiv.org/abs/2006.09306
AUTHORS:        Martin Lohmann ; Jordi Salvador ; Aniruddha Kembhavi ; Roozbeh Mottaghi
HIGHLIGHT:      Taking inspiration from infants learning from their environment through play and interaction, we present a computational framework to discover objects and learn their physical properties along this paradigm of Learning from Interaction.

63, TITLE:      Skin Segmentation from NIR Images using Unsupervised Domain Adaptation through Generative Latent Search
http://arxiv.org/abs/2006.08696
AUTHORS:        Prashant Pandey ; Aayush Kumar Tyagi ; Sameer Ambekar ; Prathosh AP
HIGHLIGHT:      We propose a method for target-independent segmentation where the 'nearest-clone' of a target image in the source domain is searched and used as a proxy in the segmentation network trained only on the source domain.

64, TITLE:      DeshuffleGAN: A Self-Supervised GAN to Improve Structure Learning
http://arxiv.org/abs/2006.08694
AUTHORS:        Gulcin Baykal ; Gozde Unal
COMMENTS:       Accepted at ICIP 2020
HIGHLIGHT:      To that end, we propose the DeshuffleGAN to enhance the learning of the discriminator and the generator, via a self-supervision approach.

65, TITLE:      Causal intersectionality for fair ranking
http://arxiv.org/abs/2006.08688
AUTHORS:        Ke Yang ; Joshua R. Loftus ; Julia Stoyanovich
HIGHLIGHT:      In this paper we propose a causal modeling approach to intersectional fairness, and a flexible, task-specific method for computing intersectionally fair rankings.

66, TITLE:      Multi-Image Summarization: Textual Summary from a Set of Cohesive Images
http://arxiv.org/abs/2006.08686
AUTHORS:        Nicholas Trieu ; Sebastian Goodman ; Pradyumna Narayana ; Kazoo Sone ; Radu Soricut
COMMENTS:       9 pages, 5 figures
HIGHLIGHT:      We propose a model that extends the image-captioning Transformer-based architecture for single image to multi-image.

67, TITLE:      To Pretrain or Not to Pretrain: Examining the Benefits of Pretraining on Resource Rich Tasks
http://arxiv.org/abs/2006.08671
AUTHORS:        Sinong Wang ; Madian Khabsa ; Hao Ma
COMMENTS:       Accepted in ACL2020
HIGHLIGHT:      This paper examines the benefits of pretrained models as a function of the number of training samples used in the downstream task.

68, TITLE:      Feature Space Saturation during Training
http://arxiv.org/abs/2006.08679
AUTHORS:        Justin Shenk ; Mats L. Richter ; Wolf Byttner ; Anders Arpteg ; Mikael Huss
COMMENTS:       23 pages, 26 figures
HIGHLIGHT:      We propose a computationally lightweight method for approximating the variance matrix during training.

69, TITLE:      A systematic review and taxonomy of explanations in decision support and recommender systems
http://arxiv.org/abs/2006.08672
AUTHORS:        Ingrid Nunes ; Dietmar Jannach
HIGHLIGHT:      In this work, we systematically review the literature on explanations in advice-giving systems.

70, TITLE:      Predicting Livelihood Indicators from Crowdsourced Street Level Images
http://arxiv.org/abs/2006.08661
AUTHORS:        Jihyeon Janel Lee ; Dylan Grosz ; Sicheng Zheng ; Burak Uzkent ; Marshall Burke ; David Lobell ; Stefano Ermon
HIGHLIGHT:      We propose an inexpensive, scalable, and interpretable approach to predict key livelihood indicators from public crowd-sourced street-level imagery.

71, TITLE:      Does it matter how well I know what you're thinking? Opponent Modelling in an RTS game
http://arxiv.org/abs/2006.08659
AUTHORS:        James Goodman ; Simon Lucas
COMMENTS:       Preprint of paper accepted for IEEE World Congress on Computational Intelligence (IEEE WCCI) 2020
HIGHLIGHT:      We investigate the sensitivity of Monte Carlo Tree Search (MCTS) and a Rolling Horizon Evolutionary Algorithm (RHEA) to the accuracy of their modelling of the opponent in a simple Real-Time Strategy game.

72, TITLE:      ESL: Entropy-guided Self-supervised Learning for Domain Adaptation in Semantic Segmentation
http://arxiv.org/abs/2006.08658
AUTHORS:        Antoine Saporta ; Tuan-Hung Vu ; Matthieu Cord ; Patrick Pérez
COMMENTS:       Accepted at the CVPR 2020 Workshop on Scalability in Autonomous Driving
HIGHLIGHT:      In this work, we propose Entropy-guided Self-supervised Learning (ESL), leveraging entropy as the confidence indicator for producing more accurate pseudo-labels.

73, TITLE:      Multiscale Deep Equilibrium Models
http://arxiv.org/abs/2006.08656
AUTHORS:        Shaojie Bai ; Vladlen Koltun ; J. Zico Kolter
HIGHLIGHT:      We propose a new class of implicit networks, the multiscale deep equilibrium model (MDEQ), suited to large-scale and highly hierarchical pattern recognition domains.

74, TITLE:      Exploiting Visual Semantic Reasoning for Video-Text Retrieval
http://arxiv.org/abs/2006.08889
AUTHORS:        Zerun Feng ; Zhimin Zeng ; Caili Guo ; Zheng Li
COMMENTS:       Accepted by IJCAI 2020. SOLE copyright holder is IJCAI (International Joint Conferences on Artificial Intelligence), all rights reserved. http://static.ijcai.org/2020-accepted_papers.html
HIGHLIGHT:      To address this issue, we propose a Visual Semantic Enhanced Reasoning Network (ViSERN) to exploit reasoning between frame regions.

75, TITLE:      The SCC-recursiveness Principle in Fuzzy Argumentation Frameworks
http://arxiv.org/abs/2006.08880
AUTHORS:        Zongshun Wang ; Jiachao Wu
HIGHLIGHT:      The SCC-recursiveness Principle in Fuzzy Argumentation Frameworks

76, TITLE:      DeepCapture: Image Spam Detection Using Deep Learning and Data Augmentation
http://arxiv.org/abs/2006.08885
AUTHORS:        Bedeuro Kim ; Sharif Abuadbba ; Hyoungshick Kim
COMMENTS:       15 pages, single column. ACISP 2020: Australasian Conference on Information Security and Privacy
HIGHLIGHT:      In this paper, we propose a new image spam email detection tool called DeepCapture using a convolutional neural network (CNN) model.

77, TITLE:      Scalable Cross Lingual Pivots to Model Pronoun Gender for Translation
http://arxiv.org/abs/2006.08881
AUTHORS:        Kellie Webster ; Emily Pitler
HIGHLIGHT:      We propose a novel cross-lingual pivoting technique for automatically producing high-quality gender labels, and show that this data can be used to fine-tune a BERT classifier with 92% F1 for Spanish dropped feminine pronouns, compared with 30-51% for neural machine translation models and 54-71% for a non-fine-tuned BERT model.

78, TITLE:      CNN Acceleration by Low-rank Approximation with Quantized Factors
http://arxiv.org/abs/2006.08878
AUTHORS:        Nikolay Kozyrskiy ; Anh-Huy Phan
HIGHLIGHT:      The greedy one-step and multi-step algorithms for the task of multilinear rank selection are proposed.

79, TITLE:      Improving accuracy and speeding up Document Image Classification through parallel systems
http://arxiv.org/abs/2006.09141
AUTHORS:        Javier Ferrando ; Juan Luis Dominguez ; Jordi Torres ; Raul Garcia ; David Garcia ; Daniel Garrido ; Jordi Cortada ; Mateo Valero
HIGHLIGHT:      This paper presents a study showing the benefits of the EfficientNet models compared with heavier Convolutional Neural Networks (CNNs) in the Document Classification task, essential problem in the digitalization process of institutions.

80, TITLE:      Cogradient Descent for Bilinear Optimization
http://arxiv.org/abs/2006.09142
AUTHORS:        Li'an Zhuo ; Baochang Zhang ; Linlin Yang ; Hanlin Chen ; Qixiang Ye ; David Doermann ; Guodong Guo ; Rongrong Ji
COMMENTS:       9 pages, 6 figures
HIGHLIGHT:      In this paper, we introduce a Cogradient Descent algorithm (CoGD) to address the bilinear problem, based on a theoretical framework to coordinate the gradient of hidden variables via a projection function.

81, TITLE:      AlphaGAN: Fully Differentiable Architecture Search for Generative Adversarial Networks
http://arxiv.org/abs/2006.09134
AUTHORS:        Yuesong Tian ; Li Shen ; Li Shen ; Guinan Su ; Zhifeng Li ; Wei Liu
HIGHLIGHT:      To this end, we propose a fully differentiable search framework for generative adversarial networks, dubbed alphaGAN.

82, TITLE:      Semantic Curiosity for Active Visual Learning
http://arxiv.org/abs/2006.09367
AUTHORS:        Devendra Singh Chaplot ; Helen Jiang ; Saurabh Gupta ; Abhinav Gupta
COMMENTS:       See project webpage at https://devendrachaplot.github.io/projects/SemanticCuriosity
HIGHLIGHT:      In this paper, we study the task of embodied interactive learning for object detection.

83, TITLE:      Evolutionary Algorithms with Self-adjusting Asymmetric Mutation
http://arxiv.org/abs/2006.09126
AUTHORS:        Amirhossein Rajabi ; Carsten Witt
COMMENTS:       16 pages. An extended abstract of this paper will be published in the proceedings of PPSN 2020
HIGHLIGHT:      Evolutionary Algorithms (EAs) and other randomized search heuristics are often considered as unbiased algorithms that are invariant with respect to different transformations of the underlying search space.

84, TITLE:      Gradient Alignment in Deep Neural Networks
http://arxiv.org/abs/2006.09128
AUTHORS:        Suraj Srinivas ; Francois Fleuret
HIGHLIGHT:      To show this, we derive a novel approximation to the score-matching objective that eliminates the need for expensive Hessian computations, which may be of independent interest.Our experiments help us identify one factor that causes input-gradient alignment in models, that being the approximate generative modelling behaviour of the normalized logit distributions.

85, TITLE:      Building One-Shot Semi-supervised (BOSS) Learning up to Fully Supervised Performance
http://arxiv.org/abs/2006.09363
AUTHORS:        Leslie N. Smith ; Adam Conovaloff
COMMENTS:       Submitted to NeurIPS 2020 conference
HIGHLIGHT:      A good prototype choice is essential and we propose a practical technique for obtaining iconic examples.

86, TITLE:      1st place solution for AVA-Kinetics Crossover in AcitivityNet Challenge 2020
http://arxiv.org/abs/2006.09116
AUTHORS:        Siyu Chen ; Junting Pan ; Guanglu Song ; Manyuan Zhang ; Hao Shao ; Ziyi Lin ; Jing Shao ; Hongsheng Li ; Yu Liu
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:2006.07976
HIGHLIGHT:      This technical report introduces our winning solution to the spatio-temporal action localization track, AVA-Kinetics Crossover, in ActivityNet Challenge 2020.

87, TITLE:      End-to-End Real-time Catheter Segmentation with Optical Flow-Guided Warping during Endovascular Intervention
http://arxiv.org/abs/2006.09117
AUTHORS:        Anh Nguyen ; Dennis Kundrat ; Giulio Dagnino ; Wenqiang Chi ; Mohamed E. M. K. Abdelaziz ; Yao Guo ; YingLiang Ma ; Trevor M. Y. Kwok ; Celia Riga ; Guang-Zhong Yang
COMMENTS:       ICRA 2020
HIGHLIGHT:      In this paper, we present FW-Net, an end-to-end and real-time deep learning framework for endovascular intervention.

88, TITLE:      How to Probe Sentence Embeddings in Low-Resource Languages: On Structural Design Choices for Probing Task Evaluation
http://arxiv.org/abs/2006.09109
AUTHORS:        Steffen Eger ; Johannes Daxenberger ; Iryna Gurevych
HIGHLIGHT:      To investigate how to probe sentence embeddings in such cases, we investigate sensitivity of probing task results to structural design choices, conducting the first such large scale study.

89, TITLE:      UCSG-Net -- Unsupervised Discovering of Constructive Solid Geometry Tree
http://arxiv.org/abs/2006.09102
AUTHORS:        Kacper Kania ; Maciej Zięba ; Tomasz Kajdanowicz
COMMENTS:       Submitted to the Thirty-fourth Conference on Neural Information Processing Systems (NeurIPS 2020), 13 pages, 7 figures
HIGHLIGHT:      On the contrary, we propose a model that extracts a CSG parse tree without any supervision - UCSG-Net.

90, TITLE:      LiDARsim: Realistic LiDAR Simulation by Leveraging the Real World
http://arxiv.org/abs/2006.09348
AUTHORS:        Sivabalan Manivasagam ; Shenlong Wang ; Kelvin Wong ; Wenyuan Zeng ; Mikita Sazanovich ; Shuhan Tan ; Bin Yang ; Wei-Chiu Ma ; Raquel Urtasun
COMMENTS:       CVPR 2020 (Oral)
HIGHLIGHT:      We tackle the problem of producing realistic simulations of LiDAR point clouds, the sensor of preference for most self-driving vehicles. To produce realistic simulations, we develop a novel simulator that captures both the power of physics-based and learning-based simulation.

91, TITLE:      Learning from Demonstration with Weakly Supervised Disentanglement
http://arxiv.org/abs/2006.09107
AUTHORS:        Yordan Hristov ; Subramanian Ramamoorthy
COMMENTS:       supplementary website at https://sites.google.com/view/weak-label-lfd
HIGHLIGHT:      We treat the task of interpretable learning from demonstration as an optimisation problem over a probabilistic generative model.

92, TITLE:      Explorable Decoding of Compressed Images
http://arxiv.org/abs/2006.09332
AUTHORS:        Yuval Bahat ; Tomer Michaeli
HIGHLIGHT:      In this work, we propose to take this idea to the realm of image decompression.

93, TITLE:      Ranking Transfer Languages with Pragmatically-Motivated Features for Multilingual Sentiment Analysis
http://arxiv.org/abs/2006.09336
AUTHORS:        Jimin Sun ; Hwijeen Ahn ; Chan Young Park ; Yulia Tsvetkov ; David R. Mortensen
HIGHLIGHT:      In this paper, we propose three pragmatically-motivated features that can help guide the optimal transfer language selection problem for cross-lingual transfer.

94, TITLE:      Progressive Skeletonization: Trimming more fat from a network at initialization
http://arxiv.org/abs/2006.09081
AUTHORS:        Pau de Jorge ; Amartya Sanyal ; Harkirat S. Behl ; Philip H. S. Torr ; Gregory Rogez ; Puneet K. Dokania
HIGHLIGHT:      To this end, we propose to find a skeletonized network with maximum foresight connection sensitivity (FORCE).

95, TITLE:      Mucko: Multi-Layer Cross-Modal Knowledge Reasoning for Fact-based VisualQuestion Answering
http://arxiv.org/abs/2006.09073
AUTHORS:        Zihao Zhu ; Jing Yu ; Yujing Wang ; Yajing Sun ; Yue Hu ; Qi Wu
COMMENTS:       Accepted by IJCAI 2020. SOLE copyright holder is IJCAI (international Joint Conferences on Artificial Intelligence)
HIGHLIGHT:      In this paper, we depict an image by a multi-modal heterogeneous graph, which contains multiple layers of information corresponding to the visual, semantic and factual features.

96, TITLE:      PERL: Pivot-based Domain Adaptation for Pre-trained Deep Contextualized Embedding Models
http://arxiv.org/abs/2006.09075
AUTHORS:        Eyal Ben-David ; Carmel Rabinovitz ; Roi Reichart
COMMENTS:       Accepted to TACL in June 2020
HIGHLIGHT:      To alleviate this, we propose PERL: A representation learning model that extends contextualized word embedding models such as BERT with pivot-based fine-tuning.

97, TITLE:      Multi-Agent Reinforcement Learning for Adaptive User Association in Dynamic mmWave Networks
http://arxiv.org/abs/2006.09066
AUTHORS:        Mohamed Sana ; Antonio De Domenico ; Wei Yu ; Yves Lostanlen ; Emilio Calvanese Strinati
COMMENTS:       Part of this work has been presented in IEEE Globecom 2019
HIGHLIGHT:      In this paper, we address this issue by designing a scalable and flexible algorithm for user association based on multi-agent reinforcement learning.

98, TITLE:      A New Run-based Connected Component Labeling for Efficiently Analyzing and Processing Holes
http://arxiv.org/abs/2006.09299
AUTHORS:        Florian Lemaitre ; Lionel Lacassagne
COMMENTS:       5 pages
HIGHLIGHT:      This article introduces a new connected component labeling and analysis algorithm for foreground and background labeling that computes the adjacency tree.

99, TITLE:      Multi-Objective CNN Based Algorithm for SAR Despeckling
http://arxiv.org/abs/2006.09050
AUTHORS:        Sergio Vitale ; Giampaolo Ferraioli ; Vito Pascazio
HIGHLIGHT:      In this paper, a convolutional neural network (CNN) with a multi-objective cost function taking care of spatial and statistical properties of the SAR image is proposed.

100, TITLE:     Model-based Adversarial Meta-Reinforcement Learning
http://arxiv.org/abs/2006.08875
AUTHORS:        Zichuan Lin ; Garrett Thomas ; Guangwen Yang ; Tengyu Ma
COMMENTS:       Code at https://github.com/LinZichuan/AdMRL
HIGHLIGHT:      We propose a minimax objective and optimize it by alternating between learning the dynamics model on a fixed task and finding the adversarial task for the current model -- the task for which the policy induced by the model is maximally suboptimal.

101, TITLE:     End-to-End Code Switching Language Models for Automatic Speech Recognition
http://arxiv.org/abs/2006.08870
AUTHORS:        Ahan M. R. ; Shreyas Sunil Kulkarni
COMMENTS:       5 pages, 2 figures, To appear in the proceedings of First Workshop on Speech Technologies for Code-switching in Multilingual Communities 2020
HIGHLIGHT:      In this paper, we particularly work on the code-switched text, one of the most common occurrences in the bilingual communities across the world.

102, TITLE:     GPU-accelerated Hierarchical Panoramic Image Feature Retrieval for Indoor Localization
http://arxiv.org/abs/2006.08861
AUTHORS:        Feng Hu
HIGHLIGHT:      This paper formulates the indoor localization problem into a multimedia retrieving problem by modeling visual landmarks with a panoramic image feature, and calculating a user's location via GPU- accelerated parallel retrieving algorithm.

103, TITLE:     Generative Semantic Hashing Enhanced via Boltzmann Machines
http://arxiv.org/abs/2006.08858
AUTHORS:        Lin Zheng ; Qinliang Su ; Dinghan Shen ; Changyou Chen
HIGHLIGHT:      In this paper, to introduce correlations among the bits of hash codes, we propose to employ the distribution of Boltzmann machine as the variational posterior.

104, TITLE:     Robust Recovery via Implicit Bias of Discrepant Learning Rates for Double Over-parameterization
http://arxiv.org/abs/2006.08857
AUTHORS:        Chong You ; Zhihui Zhu ; Qing Qu ; Yi Ma
HIGHLIGHT:      This paper shows that with a double over-parameterization for both the low-rank matrix and sparse corruption, gradient descent with discrepant learning rates provably recovers the underlying matrix even without prior knowledge on neither rank of the matrix nor sparsity of the corruption.

105, TITLE:     Grading Adjoint Logic
http://arxiv.org/abs/2006.08854
AUTHORS:        Harley Eades III ; Dominic Orchard
COMMENTS:       Extended abstract of a talk presented at LINEARITY/TLLA 2020
HIGHLIGHT:      We introduce a new logic that combines Adjoint Logic with Graded Necessity Modalities.

106, TITLE:     Optimal Sequential Task Assignment and Path Finding for Multi-Agent Robotic Assembly Planning
http://arxiv.org/abs/2006.08845
AUTHORS:        Kyle Brown ; Oriana Peltzer ; Martin A. Sehr ; Mac Schwager ; Mykel J. Kochenderfer
COMMENTS:       Presented at International Conference on Robotics and Automation (ICRA) 2020
HIGHLIGHT:      We propose a hierarchical algorithm for computing makespan-optimal solutions to the problem.

107, TITLE:     Dual-Resolution Correspondence Networks
http://arxiv.org/abs/2006.08844
AUTHORS:        Xinghui Li ; Kai Han ; Shuda Li ; Victor Adrian Prisacariu
HIGHLIGHT:      In this work, we introduce Dual-Resolution Correspondence Networks (DRC-Net), to obtain pixel-wise correspondences in a coarse-to-fine manner.

108, TITLE:     Index Selection for NoSQL Database with Deep Reinforcement Learning
http://arxiv.org/abs/2006.08842
AUTHORS:        Shun Yao ; Hongzhi Wang ; Yu Yan
HIGHLIGHT:      We propose a new approach of NoSQL database index selection.

109, TITLE:     Physics-aware Spatiotemporal Modules with Auxiliary Tasks for Meta-Learning
http://arxiv.org/abs/2006.08831
AUTHORS:        Sungyong Seo ; Chuizheng Meng ; Sirisha Rambhatla ; Yan Liu
HIGHLIGHT:      In this paper, we propose a framework, physics-aware modular meta-learning with auxiliary tasks (PiMetaL) whose spatial modules incorporate PDE-independent knowledge and temporal modules are rapidly adaptable to the limited data, respectively.

110, TITLE:     Explainable AI for a No-Teardown Vehicle Component Cost Estimation: A Top-Down Approach
http://arxiv.org/abs/2006.08828
AUTHORS:        Ayman Moawad ; Ehsan Islam ; Namdoo Kim ; Ram Vijayagopal ; Aymeric Rousseau ; Wei Biao Wu
COMMENTS:       17 pages, 18 figures
HIGHLIGHT:      Particularly, we present a data-driven approach to vehicle price modeling and its component price estimation by leveraging a combination of concepts from machine learning and game theory.

111, TITLE:     Cardiac Segmentation with Strong Anatomical Guarantees
http://arxiv.org/abs/2006.08825
AUTHORS:        Nathan Painchaud ; Youssef Skandarani ; Thierry Judge ; Olivier Bernard ; Alain Lalande ; Pierre-Marc Jodoin
COMMENTS:       11 pages, accepted for publication in IEEE TMI
HIGHLIGHT:      In this paper, we present a framework for producing cardiac image segmentation maps that are guaranteed to respect pre-defined anatomical criteria, while remaining within the inter-expert variability.

112, TITLE:     Quantitatively Assessing the Benefits of Model-driven Development in Agent-based Modeling and Simulation
http://arxiv.org/abs/2006.08820
AUTHORS:        Fernando Santos ; Ingrid Nunes ; Ana L. C. Bazzan
HIGHLIGHT:      We thus in this paper present an empirical study that quantitatively compares the use of MDD and ABMS platforms mainly in terms of effort and developer mistakes.

113, TITLE:     A study of the effect of the illumination model on the generation of synthetic training datasets
http://arxiv.org/abs/2006.08819
AUTHORS:        Xin Zhang ; Ning Jia ; Ioannis Ivrissimtzis
COMMENTS:       8 pages
HIGHLIGHT:      In this paper, we study how the illumination model used by the rendering software affects the quality of the generated images. We created eight training sets, each one with a different illumination model, and tested them on three different network architectures, ResNet, U-Net and a combined architecture developed by us.

114, TITLE:     Explaining reputation assessments
http://arxiv.org/abs/2006.08818
AUTHORS:        Ingrid Nunes ; Phillip Taylor ; Lina Barakat ; Nathan Griffiths ; Simon Miles
HIGHLIGHT:      In this paper, we propose an approach to explain the rationale behind assessments from quantitative reputation models, by generating arguments that are combined to form explanations.

115, TITLE:     A Particle Swarm Optimization hyper-heuristic for the Dynamic Vehicle Routing Problem
http://arxiv.org/abs/2006.08809
AUTHORS:        Michał Okulewicz ; Jacek Mańdziuk
COMMENTS:       14 pages, presented at BIOMA 2016 conference, Bled, Slovenia
HIGHLIGHT:      This paper presents a method for choosing a Particle Swarm Optimization based optimizer for the Dynamic Vehicle Routing Problem on the basis of the initially available data of a given problem instance.



==========Updates to Previous Papers==========
1, TITLE:       Multi-Modal Fingerprint Presentation Attack Detection: Evaluation On A New Dataset
http://arxiv.org/abs/2006.07498
AUTHORS:        Leonidas Spinoulas ; Hengameh Mirzaalian ; Mohamed Hussein ; Wael AbdAlmageed
HIGHLIGHT:      In this work, rather than relying on legacy fingerprint images, which are widely used in the community, we study the usefulness of multiple recently introduced sensing modalities.

2, TITLE:       Towards Deployment of Robust AI Agents for Human-Machine Partnerships
http://arxiv.org/abs/1910.02330
AUTHORS:        Ahana Ghosh ; Sebastian Tschiatschek ; Hamed Mahdavi ; Adish Singla
HIGHLIGHT:      We study the problem of designing AI agents that can robustly cooperate with people in human-machine partnerships.

3, 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:       20 pages, 22 figures, under review
HIGHLIGHT:      In this paper, we propose a comprehensive study of these problems in the light of the multi-dimensional histograms of deep features.

4, TITLE:       LayoutLM: Pre-training of Text and Layout for Document Image Understanding
http://arxiv.org/abs/1912.13318
AUTHORS:        Yiheng Xu ; Minghao Li ; Lei Cui ; Shaohan Huang ; Furu Wei ; Ming Zhou
COMMENTS:       KDD 2020
HIGHLIGHT:      In this paper, we propose the \textbf{LayoutLM} to jointly model interactions between text and layout information across scanned document images, which is beneficial for a great number of real-world document image understanding tasks such as information extraction from scanned documents.

5, TITLE:       Effectively Unbiased FID and Inception Score and where to find them
http://arxiv.org/abs/1911.07023
AUTHORS:        Min Jin Chong ; David Forsyth
COMMENTS:       CVPR 2020
HIGHLIGHT:      This paper shows that two commonly used evaluation metrics for generative models, the Fr\'echet Inception Distance (FID) and the Inception Score (IS), are biased -- the expected value of the score computed for a finite sample set is not the true value of the score.

6, TITLE:       Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion
http://arxiv.org/abs/1912.08795
AUTHORS:        Hongxu Yin ; Pavlo Molchanov ; Zhizhong Li ; Jose M. Alvarez ; Arun Mallya ; Derek Hoiem ; Niraj K. Jha ; Jan Kautz
HIGHLIGHT:      We introduce DeepInversion, a new method for synthesizing images from the image distribution used to train a deep neural network.

7, TITLE:       ProbAct: A Probabilistic Activation Function for Deep Neural Networks
http://arxiv.org/abs/1905.10761
AUTHORS:        Kumar Shridhar ; Joonho Lee ; Hideaki Hayashi ; Purvanshi Mehta ; Brian Kenji Iwana ; Seokjun Kang ; Seiichi Uchida ; Sheraz Ahmed ; Andreas Dengel
HIGHLIGHT:      In this work, we propose a novel probabilistic activation function, called ProbAct.

8, TITLE:       Dilated Convolutions with Lateral Inhibitions for Semantic Image Segmentation
http://arxiv.org/abs/2006.03708
AUTHORS:        Yujiang Wang ; Mingzhi Dong ; Jie Shen ; Yiming Lin ; Maja Pantic
HIGHLIGHT:      Inspired by the Lateral Inhibition (LI) mechanisms in human visual systems, we propose the dilated convolution with lateral inhibitions (LI-Convs) to overcome these limitations.

9, TITLE:       AutoGrow: Automatic Layer Growing in Deep Convolutional Networks
http://arxiv.org/abs/1906.02909
AUTHORS:        Wei Wen ; Feng Yan ; Yiran Chen ; Hai Li
COMMENTS:       KDD 2020
HIGHLIGHT:      We propose AutoGrow to automate depth discovery in DNNs: starting from a shallow seed architecture, AutoGrow grows new layers if the growth improves the accuracy; otherwise, stops growing and thus discovers the depth.

10, TITLE:      Information Extraction of Clinical Trial Eligibility Criteria
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(dot)gov to a shared knowledge base.

11, TITLE:      StickyPillars: Robust and Efficient Feature Matching on Point Clouds using Graph Neural Networks
http://arxiv.org/abs/2002.03983
AUTHORS:        Martin Simon ; Kai Fischer ; Stefan Milz ; Christian Witt ; Florian Oelsner ; Patrick Maeder ; Horst-Michael Gross
HIGHLIGHT:      We overcome these drawbacks by introducing StickyPillars, an end-to-end trained 3D feature matching approach based on a graph neural network.

12, TITLE:      Pixel Invisibility: Detecting Objects Invisible in Color Images
http://arxiv.org/abs/2006.08383
AUTHORS:        Yongxin Wang ; Duminda Wijesekera
COMMENTS:       8 pages, 7 figures, submitted to NIPS 2020
HIGHLIGHT:      We propose a novel use of cross modal knowledge distillation from color to infra-red domain using weakly-aligned image pairs from the day and construct indicators for the pixel-level invisibility based on the distances of their intermediate-level features.

13, TITLE:      Reward-rational (implicit) choice: A unifying formalism for reward learning
http://arxiv.org/abs/2002.04833
AUTHORS:        Hong Jun Jeon ; Smitha Milli ; Anca D. Dragan
HIGHLIGHT:      Our key insight is that different types of behavior can be interpreted in a single unifying formalism - as a reward-rational choice that the human is making, often implicitly.

14, TITLE:      Look Locally Infer Globally: A Generalizable Face Anti-Spoofing Approach
http://arxiv.org/abs/2006.02834
AUTHORS:        Debayan Deb ; Anil K. Jain
HIGHLIGHT:      Given that face anti-spoofing is inherently a local task, we propose a face anti-spoofing framework, namely Self-Supervised Regional Fully Convolutional Network (SSR-FCN), that is trained to learn local discriminative cues from a face image in a self-supervised manner.

15, TITLE:      Learning to Forget for Meta-Learning
http://arxiv.org/abs/1906.05895
AUTHORS:        Sungyong Baik ; Seokil Hong ; Kyoung Mu Lee
COMMENTS:       CVPR 2020. Code at https://github.com/baiksung/L2F
HIGHLIGHT:      Thus, we propose task-and-layer-wise attenuation on the compromised initialization to reduce its influence.

16, TITLE:      Integrating Temporal Information to Spatial Information in a Neural Circuit
http://arxiv.org/abs/1903.01217
AUTHORS:        Nancy Lynch ; Mien Brabeeba Wang
HIGHLIGHT:      In this paper, we consider networks of deterministic spiking neurons, firing synchronously at discrete times; such spiking neural networks are inspired by networks of neurons and synapses that occur in brains.

17, TITLE:      Hold me tight! Influence of discriminative features on deep network boundaries
http://arxiv.org/abs/2002.06349
AUTHORS:        Guillermo Ortiz-Jimenez ; Apostolos Modas ; Seyed-Mohsen Moosavi-Dezfooli ; Pascal Frossard
HIGHLIGHT:      In this work, we borrow tools from the field of adversarial robustness, and propose a new perspective that relates dataset features to the distance of samples to the decision boundary.

18, TITLE:      Learning 3D-3D Correspondences for One-shot Partial-to-partial Registration
http://arxiv.org/abs/2006.04523
AUTHORS:        Zheng Dang ; Fei Wang ; Mathieu Salzmann
COMMENTS:       11 pages
HIGHLIGHT:      To this end, we propose an Optimal Transport layer able to account for occluded points thanks to the use of outlier bins.

19, TITLE:      SQuINTing at VQA Models: Introspecting VQA Models with Sub-Questions
http://arxiv.org/abs/2001.06927
AUTHORS:        Ramprasaath R. Selvaraju ; Purva Tendulkar ; Devi Parikh ; Eric Horvitz ; Marco Ribeiro ; Besmira Nushi ; Ece Kamar
COMMENTS:       Accepted to CVPR'20 as an Oral Presentation
HIGHLIGHT:      To address this shortcoming, we propose an approach called Sub-Question Importance-aware Network Tuning (SQuINT), which encourages the model to attend to the same parts of the image when answering the reasoning question and the perception sub question.

20, TITLE:      Visual Transformers: Token-based Image Representation and Processing for Computer Vision
http://arxiv.org/abs/2006.03677
AUTHORS:        Bichen Wu ; Chenfeng Xu ; Xiaoliang Dai ; Alvin Wan ; Peizhao Zhang ; Masayoshi Tomizuka ; Kurt Keutzer ; Peter Vajda
HIGHLIGHT:      In this work, we challenge this paradigm: we instead (a) represent images as a set of visual tokens and (b) apply visual transformers to find relationships between visual semantic concepts.

21, TITLE:      Bayesian Reasoning with Deep-Learned Knowledge
http://arxiv.org/abs/2001.11031
AUTHORS:        Jakob Knollmüller ; Torsten Enßlin
COMMENTS:       11 pages, 5 figures
HIGHLIGHT:      We use independently trained neural networks to represent abstract concepts and combine them through Bayesian reasoning to approach tasks outside their initial scope.

22, TITLE:      The Impact of Non-stationarity on Generalisation in Deep Reinforcement Learning
http://arxiv.org/abs/2006.05826
AUTHORS:        Maximilian Igl ; Gregory Farquhar ; Jelena Luketina ; Wendelin Boehmer ; Shimon Whiteson
HIGHLIGHT:      Consequently, to improve generalisation of deep RL agents, we propose Iterated Relearning (ITER).

23, TITLE:      Training BatchNorm and Only BatchNorm: On the Expressive Power of Random Features in CNNs
http://arxiv.org/abs/2003.00152
AUTHORS:        Jonathan Frankle ; David J. Schwab ; Ari S. Morcos
COMMENTS:       NeurIPS submission
HIGHLIGHT:      To study this question, we investigate the performance achieved when training only these parameters and freezing all others at their random initializations.

24, TITLE:      Measuring Forecasting Skill from Text
http://arxiv.org/abs/2006.07425
AUTHORS:        Shi Zong ; Alan Ritter ; Eduard Hovy
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      In this paper we explore connections between the language people use to describe their predictions and their forecasting skill.

25, TITLE:      Effective writing style imitation via combinatorial paraphrasing
http://arxiv.org/abs/1905.13464
AUTHORS:        Tommi Gröndahl ; N. Asokan
COMMENTS:       16 pages, 1 figure, Accepted for publication in Privacy Enhancing Technologies (PETS2020)
HIGHLIGHT:      To mitigate this problem we propose ParChoice: a technique based on the combinatorial application of multiple paraphrasing algorithms.

26, TITLE:      Learning Individually Fair Classifier with Path-Specific Causal-Effect Constraint
http://arxiv.org/abs/2002.06746
AUTHORS:        Yoichi Chikahara ; Shinsaku Sakaue ; Akinori Fujino ; Hisashi Kashima
COMMENTS:       28 pages, 8 figures, 4 tables
HIGHLIGHT:      In this paper, we propose a framework for learning an individually fair classifier without relying on the causal model.

27, TITLE:      Toward Adversarial Robustness via Semi-supervised Robust Training
http://arxiv.org/abs/2003.06974
AUTHORS:        Yiming Li ; Baoyuan Wu ; Yan Feng ; Yanbo Fan ; Yong Jiang ; Zhifeng Li ; Shutao Xia
COMMENTS:       19 pages
HIGHLIGHT:      In this work, we propose a novel defense method, the robust training (RT), by jointly minimizing two separated risks ($R_{stand}$ and $R_{rob}$), which is with respect to the benign example and its neighborhoods respectively.

28, TITLE:      Clean-Label Backdoor Attacks on Video Recognition Models
http://arxiv.org/abs/2003.03030
AUTHORS:        Shihao Zhao ; Xingjun Ma ; Xiang Zheng ; James Bailey ; Jingjing Chen ; Yu-Gang Jiang
COMMENTS:       CVPR2020
HIGHLIGHT:      In this paper, we show that existing image backdoor attacks are far less effective on videos, and outline 4 strict conditions where existing attacks are likely to fail: 1) scenarios with more input dimensions (eg.

29, TITLE:      Continual General Chunking Problem and SyncMap
http://arxiv.org/abs/2006.07853
AUTHORS:        Danilo Vasconcellos Vargas ; Toshitake Asabuki
HIGHLIGHT:      Here, we propose a continual generalization of the chunking problem (an unsupervised problem), encompassing fixed and probabilistic chunks, discovery of temporal and causal structures and their continual variations.

30, TITLE:      FP-Stereo: Hardware-Efficient Stereo Vision for Embedded Applications
http://arxiv.org/abs/2006.03250
AUTHORS:        Jieru Zhao ; Tingyuan Liang ; Liang Feng ; Wenchao Ding ; Sharad Sinha ; Wei Zhang ; Shaojie Shen
COMMENTS:       IEEE International Conference on Field Programmable Logic and Applications (FPL), 2020
HIGHLIGHT:      To reduce the design effort and achieve the right balance, we propose FP-Stereo for building high-performance stereo matching pipelines on FPGAs automatically.

31, TITLE:      Catch & Carry: Reusable Neural Controllers for Vision-Guided Whole-Body Tasks
http://arxiv.org/abs/1911.06636
AUTHORS:        Josh Merel ; Saran Tunyasuvunakool ; Arun Ahuja ; Yuval Tassa ; Leonard Hasenclever ; Vu Pham ; Tom Erez ; Greg Wayne ; Nicolas Heess
HIGHLIGHT:      We address the longstanding challenge of producing flexible, realistic humanoid character controllers that can perform diverse whole-body tasks involving object interactions.

32, TITLE:      Few-shot Object Detection on Remote Sensing Images
http://arxiv.org/abs/2006.07826
AUTHORS:        Jingyu Deng ; Xiang Li ; Yi Fang
COMMENTS:       12pages, 7 figures
HIGHLIGHT:      In this paper, we deal with the problem of object detection on remote sensing images.

33, TITLE:      Improved algorithm for permutation testing
http://arxiv.org/abs/2006.08473
AUTHORS:        Xiaojin Zhang
HIGHLIGHT:      In this paper, we provide a simple adaptive algorithm with one-sided error for testing monotone permutation.

34, TITLE:      Prototype Rectification for Few-Shot Learning
http://arxiv.org/abs/1911.10713
AUTHORS:        Jinlu Liu ; Liang Song ; Yongqiang Qin
HIGHLIGHT:      In this paper, we figure out two key influencing factors of the process: the intra-class bias and the cross-class bias.

35, TITLE:      Graph Neural Ordinary Differential Equations
http://arxiv.org/abs/1911.07532
AUTHORS:        Michael Poli ; Stefano Massaroli ; Junyoung Park ; Atsushi Yamashita ; Hajime Asama ; Jinkyoo Park
HIGHLIGHT:      We introduce the framework of continuous--depth graph neural networks (GNNs).

36, TITLE:      Image Restoration from Parametric Transformations using Generative Models
http://arxiv.org/abs/2005.14036
AUTHORS:        Kalliopi Basioti ; George V. Moustakides
HIGHLIGHT:      When images are statistically described by a generative model we can use this information to develop optimum techniques for various image restoration problems as inpainting, super-resolution, image coloring, generative model inversion, etc.

37, TITLE:      NTIRE 2020 Challenge on Video Quality Mapping: Methods and Results
http://arxiv.org/abs/2005.02291
AUTHORS:        Dario Fuoli ; Zhiwu Huang ; Martin Danelljan ; Radu Timofte ; Hua Wang ; Longcun Jin ; Dewei Su ; Jing Liu ; Jaehoon Lee ; Michal Kudelski ; Lukasz Bala ; Dmitry Hrybov ; Marcin Mozejko ; Muchen Li ; Siyao Li ; Bo Pang ; Cewu Lu ; Chao Li ; Dongliang He ; Fu Li ; Shilei Wen
COMMENTS:       The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops
HIGHLIGHT:      This paper reviews the NTIRE 2020 challenge on video quality mapping (VQM), which addresses the issues of quality mapping from source video domain to target video domain.

38, TITLE:      New Results for the Complexity of Resilience for Binary Conjunctive Queries with Self-Joins
http://arxiv.org/abs/1907.01129
AUTHORS:        Cibele Freire ; Wolfgang Gatterbauer ; Neil Immerman ; Alexandra Meliou
COMMENTS:       23 pages, 19 figures, included a new section
HIGHLIGHT:      In this paper, we give several novel results on the hardness of the resilience problem for $\textit{binary conjunctive queries with self-joins}$ (i.e. conjunctive queries with relations of maximal arity 2) with one repeated relation.

39, TITLE:      Mining Implicit Relevance Feedback from User Behavior for Web Question Answering
http://arxiv.org/abs/2006.07581
AUTHORS:        Linjun Shou ; Shining Bo ; Feixiang Cheng ; Ming Gong ; Jian Pei ; Daxin Jiang
COMMENTS:       Accepted by KDD 2020
HIGHLIGHT:      In this paper, we make the first study to explore the correlation between user behavior and passage relevance, and propose a novel approach for mining training data for Web QA.

40, TITLE:      Addressing target shift in zero-shot learning using grouped adversarial learning
http://arxiv.org/abs/2003.00845
AUTHORS:        Saneem Ahmed Chemmengath ; Soumava Paul ; Samarth Bharadwaj ; Suranjana Samanta ; Karthik Sankaranarayanan
COMMENTS:       Under submission at Neurips 2020
HIGHLIGHT:      In this paper, we present a new paradigm for ZSL that: (i) utilizes the class-attribute mapping of unseen classes to estimate the change in target distribution (target shift), and (ii) propose a novel technique called grouped Adversarial Learning (gAL) to reduce negative effects of this shift.

41, TITLE:      Deep Multi-View Enhancement Hashing for Image Retrieval
http://arxiv.org/abs/2002.00169
AUTHORS:        Chenggang Yan ; Biao Gong ; Yuxuan Wei ; Yue Gao
HIGHLIGHT:      In this paper, we propose a supervised multi-view hash model which can enhance the multi-view information through neural networks.

42, TITLE:      Morphing Attack Detection -- Database, Evaluation Platform and Benchmarking
http://arxiv.org/abs/2006.06458
AUTHORS:        Kiran Raja ; Matteo Ferrara ; Annalisa Franco ; Luuk Spreeuwers ; Illias Batskos ; Florens de Wit Marta Gomez-Barrero ; Ulrich Scherhag ; Daniel Fischer ; Sushma Venkatesh ; Jag Mohan Singh ; Guoqiang Li ; Loïc Bergeron ; Sergey Isadskiy ; Raghavendra Ramachandra ; Christian Rathgeb ; Dinusha Frings ; Uwe Seidel ; Fons Knopjes ; Raymond Veldhuis ; Davide Maltoni ; Christoph Busch
COMMENTS:       The following paper is a pre-print. The publication is currently under review for IEEE Transactions on Information Forensics and Security (TIFS)
HIGHLIGHT:      In this work, we present a new sequestered dataset for facilitating the advancements of MAD where the algorithms can be tested on unseen data in an effort to better generalize.

43, TITLE:      Optimality and limitations of audio-visual integration for cognitive systems
http://arxiv.org/abs/1912.00581
AUTHORS:        W. Paul Boyce ; Tony Lindsay ; Arkady Zgonnikov ; Ignacio Rano ; KongFatt Wong-Lin
COMMENTS:       20 pages, 6 figures, 1 table 16/06/2020: Updated version includes expanded discussion and addition of new references. Also updated author affiliation information. This version has been accepted for publication with Frontiers
HIGHLIGHT:      We review audio-visual facilitations and illusions that are products of multisensory integration, and the computational models that account for these phenomena.

44, TITLE:      Deformable 3D Convolution for Video Super-Resolution
http://arxiv.org/abs/2004.02803
AUTHORS:        Xinyi Ying ; Longguang Wang ; Yingqian Wang ; Weidong Sheng ; Wei An ; Yulan Guo
COMMENTS:       Submitted to IEEE Signal Processing Letters as a revised version. Code is available at: https://github.com/XinyiYing/D3Dnet. A demo video can be viewed at: https://wyqdatabase.s3-us-west-1.amazonaws.com/D3Dnet.mp4
HIGHLIGHT:      In this paper, we propose a deformable 3D convolution network (D3Dnet) to incorporate spatio-temporal information from both spatial and temporal dimensions for video SR.

45, TITLE:      Ultra Fast Structure-aware Deep Lane Detection
http://arxiv.org/abs/2004.11757
AUTHORS:        Zequn Qin ; Huanyu Wang ; Xi Li
HIGHLIGHT:      Motivated by this observation, we propose a novel, simple, yet effective formulation aiming at extremely fast speed and challenging scenarios.

46, TITLE:      Harnessing Code Switching to Transcend the Linguistic Barrier
http://arxiv.org/abs/2001.11258
AUTHORS:        Ashiqur R. KhudaBukhsh ; Shriphani Palakodety ; Jaime G. Carbonell
HIGHLIGHT:      In this paper, we provide a systematic approach to sample code mixed documents leveraging a polyglot embedding based method that requires minimal supervision.

47, TITLE:      Disentangling Image Distortions in Deep Feature Space
http://arxiv.org/abs/2002.11409
AUTHORS:        Simone Bianco ; Luigi Celona ; Paolo Napoletano
HIGHLIGHT:      In this work we take a further step in the direction of a broader understanding of such property by analyzing the capability of deep visual representations to intrinsically characterize different types of image distortions.

48, TITLE:      A Proposal for a Revision of ISO Modula-2
http://arxiv.org/abs/2006.07193
AUTHORS:        Benjamin Kowarsch
COMMENTS:       Note: This paper contains the same *terms of reference* as arXiv:1809.07080 (a different paper by the same author) which is erroneously marked as duplication by arXiv's automated process. Changes: This version adds several footnotes and two more references, some formatting has been modified, content remains the same as v1
HIGHLIGHT:      This paper discusses some of the deficiencies of IS 10514-1 and proposes a limited revision that could be carried out with moderate effort.

49, TITLE:      UCLID-Net: Single View Reconstruction in Object Space
http://arxiv.org/abs/2006.03817
AUTHORS:        Benoit Guillard ; Edoardo Remelli ; Pascal Fua
COMMENTS:       Added supplementary material
HIGHLIGHT:      In this paper, we show that building a geometry preserving 3-dimensional latent space helps the network concurrently learn global shape regularities and local reasoning in the object coordinate space and, as a result, boosts performance.

50, TITLE:      Visual-Semantic Graph Attention Networks for Human-Object Interaction Detection
http://arxiv.org/abs/2001.02302
AUTHORS:        Zhijun Liang ; Junfa Liu ; Yisheng Guan ; Juan Rojas
COMMENTS:       Update the results on HICO-DET and V-COCO dataset. 10 pages, 4 figures, 2 tables
HIGHLIGHT:      We contribute a dual-graph attention network that effectively aggregates contextual visual, spatial, and semantic information dynamically from primary subject-object relations as well as subsidiary relations through attention mechanisms for strong disambiguating power.

51, TITLE:      The PSPACE-hardness of understanding neural circuits
http://arxiv.org/abs/2006.08266
AUTHORS:        Vidya Sagar Sharma ; Piyush Srivastava
COMMENTS:       2 figures
HIGHLIGHT:      In this paper, we prove that the problems of finding minimal or minimum-size degenerate sets, and of finding the set of vital neurons, of a neural circuit given as input, are in fact PSPACE-hard.

52, TITLE:      Effective and Efficient Computation with Multiple-timescale Spiking Recurrent Neural Networks
http://arxiv.org/abs/2005.11633
AUTHORS:        Bojian Yin ; Federico Corradi ; Sander M. Bohté
COMMENTS:       11 pages,5 figures
HIGHLIGHT:      From this, we calculate a $>$100x energy improvement for our SRNNs over classical RNNs on the harder tasks.

53, TITLE:      The DeepFake Detection Challenge Dataset
http://arxiv.org/abs/2006.07397
AUTHORS:        Brian Dolhansky ; Joanna Bitton ; Ben Pflaum ; Jikuo Lu ; Russ Howes ; Menglin Wang ; Cristian Canton Ferrer
HIGHLIGHT:      To counter this emerging threat, we have constructed an extremely large face swap video dataset to enable the training of detection models, and organized the accompanying DeepFake Detection Challenge (DFDC) Kaggle competition.

54, TITLE:      Going Deeper with Lean Point Networks
http://arxiv.org/abs/1907.00960
AUTHORS:        Eric-Tuan Le ; Iasonas Kokkinos ; Niloy J. Mitra
COMMENTS:       16 pages, 11 figures, 9 tables
HIGHLIGHT:      In this work we introduce Lean Point Networks (LPNs) to train deeper and more accurate point processing networks by relying on three novel point processing blocks that improve memory consumption, inference time, and accuracy: a convolution-type block for point sets that blends neighborhood information in a memory-efficient manner; a crosslink block that efficiently shares information across low- and high-resolution processing branches; and a multiresolution point cloud processing block for faster diffusion of information.

55, TITLE:      MC-BERT: Efficient Language Pre-Training via a Meta Controller
http://arxiv.org/abs/2006.05744
AUTHORS:        Zhenhui Xu ; Linyuan Gong ; Guolin Ke ; Di He ; Shuxin Zheng ; Liwei Wang ; Jiang Bian ; Tie-Yan Liu
HIGHLIGHT:      To achieve better efficiency and effectiveness, we propose a novel meta-learning framework, MC-BERT.

56, 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).

57, TITLE:      A Bayesian Inference Framework for Procedural Material Parameter Estimation
http://arxiv.org/abs/1912.01067
AUTHORS:        Yu Guo ; Milos Hasan ; Lingqi Yan ; Shuang Zhao
COMMENTS:       10 pages, 6 figures
HIGHLIGHT:      We explore the inverse rendering problem of procedural material parameter estimation from photographs, presenting a unified view of the problem in a Bayesian framework.

58, TITLE:      Leveraging Multimodal Behavioral Analytics for Automated Job Interview Performance Assessment and Feedback
http://arxiv.org/abs/2006.07909
AUTHORS:        Anumeha Agrawal ; Rosa Anil George ; Selvan Sunitha Ravi ; Sowmya Kamath S ; Anand Kumar M
COMMENTS:       9 pages, ACL 2020
HIGHLIGHT:      We propose a multimodal analytical framework that analyzes the candidate in an interview scenario and provides feedback for predefined labels such as engagement, speaking rate, eye contact, etc.

59, TITLE:      Bridging the Gap between Spatial and Spectral Domains: A Survey on Graph Neural Networks
http://arxiv.org/abs/2002.11867
AUTHORS:        Zhiqian Chen ; Fanglan Chen ; Lei Zhang ; Taoran Ji ; Kaiqun Fu ; Liang Zhao ; Feng Chen ; Chang-Tien Lu
HIGHLIGHT:      This paper proposes a unified framework and provides a novel perspective that can widely fit existing GNNs into our framework methodologically.

60, TITLE:      Hierarchical Neural Architecture Search for Single Image Super-Resolution
http://arxiv.org/abs/2003.04619
AUTHORS:        Yong Guo ; Yongsheng Luo ; Zhenhao He ; Jin Huang ; Jian Chen
COMMENTS:       This paper is accepted by IEEE Signal Processing Letters
HIGHLIGHT:      To address the above issues, we propose a Hierarchical Neural Architecture Search (HNAS) method to automatically design promising architectures with different requirements of computation cost.

61, TITLE:      On Defending Against Label Flipping Attacks on Malware Detection Systems
http://arxiv.org/abs/1908.04473
AUTHORS:        Rahim Taheri ; Reza Javidan ; Mohammad Shojafar ; Zahra Pooranian ; Ali Miri ; Mauro Conti
COMMENTS:       21 pages, 6 figures, 4 tables, NCAA Springer Journal
HIGHLIGHT:      In this paper, we design an architecture to tackle the Android malware detection problem in IoT systems.

62, TITLE:      Automatic Creation of Text Corpora for Low-Resource Languages from the Internet: The Case of Swiss German
http://arxiv.org/abs/1912.00159
AUTHORS:        Lucy Linder ; Michael Jungo ; Jean Hennebert ; Claudiu Musat ; Andreas Fischer
HIGHLIGHT:      This paper presents SwissCrawl, the largest Swiss German text corpus to date.

63, TITLE:      A Journey into Ontology Approximation: From Non-Horn to Horn
http://arxiv.org/abs/2001.07754
AUTHORS:        Anneke Haga ; Carsten Lutz ; Johannes Marti ; Frank Wolter
COMMENTS:       21 pages, 4 figures, paragraph with examples added
HIGHLIGHT:      We study complete approximations of an ontology formulated in a non-Horn description logic (DL) such as $\mathcal{ALC}$ in a Horn DL such as~$\mathcal{EL}$.

64, TITLE:      Probing Neural Language Models for Human Tacit Assumptions
http://arxiv.org/abs/2004.04877
AUTHORS:        Nathaniel Weir ; Adam Poliak ; Benjamin Van Durme
COMMENTS:       To be published in CogSci 2020
HIGHLIGHT:      We find models to be profoundly effective at retrieving concepts given associated properties.

65, TITLE:      Dual Learning for Semi-Supervised Natural Language Understanding
http://arxiv.org/abs/2004.12299
AUTHORS:        Su Zhu ; Ruisheng Cao ; Kai Yu
COMMENTS:       12 pages, 4 figures; Accepted for IEEE/ACM Transactions on Audio Speech and Language Processing
HIGHLIGHT:      In this work, we introduce a dual task of NLU, semantic-to-sentence generation (SSG), and propose a new framework for semi-supervised NLU with the corresponding dual model.

66, TITLE:      Folding-based compression of point cloud attributes
http://arxiv.org/abs/2002.04439
AUTHORS:        Maurice Quach ; Giuseppe Valenzise ; Frederic Dufaux
COMMENTS:       To be published in ICIP 2020
HIGHLIGHT:      In this work, we consider point cloud attribute compression; thus, we compress this image with a conventional 2D image codec.

67, TITLE:      Statistical Analysis on E-Commerce Reviews, with Sentiment Classification using Bidirectional Recurrent Neural Network (RNN)
http://arxiv.org/abs/1805.03687
AUTHORS:        Abien Fred Agarap
COMMENTS:       21 pages, 29 figures, 6 tables
HIGHLIGHT:      This paper attempts to understand the correlation of different variables in customer reviews on a women clothing e-commerce, and to classify each review whether it recommends the reviewed product or not and whether it consists of positive, negative, or neutral sentiment.

68, TITLE:      Relational reasoning and generalization using non-symbolic neural networks
http://arxiv.org/abs/2006.07968
AUTHORS:        Atticus Geiger ; Alexandra Carstensen ; Michael C. Frank ; Christopher Potts
HIGHLIGHT:      This paper revisits the question of whether equality can be learned by neural networks that do not encode explicit symbolic structure.

69, TITLE:      Exploring Weaknesses of VQA Models through Attribution Driven Insights
http://arxiv.org/abs/2006.06637
AUTHORS:        Shaunak Halbe
COMMENTS:       Second Grand-Challenge and Workshop on Multimodal Language, ACL 2020
HIGHLIGHT:      Recent research effectively applies these VQA models for answering visual questions for the blind.

70, TITLE:      Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue
http://arxiv.org/abs/2002.07510
AUTHORS:        Byeongchang Kim ; Jaewoo Ahn ; Gunhee Kim
COMMENTS:       Published in ICLR 2020 (Spotlight)
HIGHLIGHT:      As we focus on better modeling the knowledge selection in the multi-turn knowledge-grounded dialogue, we propose a sequential latent variable model as the first approach to this matter.

71, TITLE:      Learning sums of powers of low-degree polynomials in the non-degenerate case
http://arxiv.org/abs/2004.06898
AUTHORS:        Ankit Garg ; Neeraj Kayal ; Chandan Saha
COMMENTS:       Fixed a minor bug in the statement and proof of Corollary 3.1
HIGHLIGHT:      In this paper, we give a $\text{poly}((ns)^t)$-time learning algorithm for finding the $Q_i$'s given (black-box access to) $f$, if the $Q_i's$ satisfy certain non-degeneracy conditions and $n$ is larger than $d^2$.

72, 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.

73, TITLE:      A Category Theory Approach to Interoperability
http://arxiv.org/abs/2005.06872
AUTHORS:        Riccardo Del Gratta
COMMENTS:       Paper submitted to Applied Category Theory 2020 and accepted for Virtual Poster Session
HIGHLIGHT:      In this article, we propose a Category Theory approach to (syntactic) interoperability between linguistic tools.