2020.06.09.txt

==========New Papers==========
1, TITLE:       Finger Texture Biometric Characteristic: a Survey
http://arxiv.org/abs/2006.04193
AUTHORS:        Raid R. O. Al-Nima ; Tingting Han ; Taolue Chen ; Satnam Dlay ; Jonathon Chambers
HIGHLIGHT:      In this paper, a comprehensive survey of the relevant FT studies is presented.

2, TITLE:       Predictive Coding Approximates Backprop along Arbitrary Computation Graphs
http://arxiv.org/abs/2006.04182
AUTHORS:        Beren Millidge ; Alexander Tschantz ; Christopher L. Buckley
COMMENTS:       Submitted to NeurIPS 2020
HIGHLIGHT:      Here, we demonstrate that predictive coding converges asymptotically (and in practice rapidly) to exact backprop gradients on arbitrary computation graphs using only local learning rules.

3, TITLE:       Realistic text replacement with non-uniform style conditioning
http://arxiv.org/abs/2006.04170
AUTHORS:        Arseny Nerinovsky ; Igor Buzhinsky ; Andey Filchencov
COMMENTS:       8 pages, 5 figures
HIGHLIGHT:      In this work, we study the possibility of realistic text replacement, the goal of which is to replace text present in the image with user-supplied text.

4, TITLE:       Learning pose variations within shape population by constrained mixtures of factor analyzers
http://arxiv.org/abs/2006.04171
AUTHORS:        Xilu Wang
COMMENTS:       25 Pages, 15 Figures
HIGHLIGHT:      This paper formulates the pose learning problem as mixtures of factor analyzers.

5, TITLE:       Deep active inference agents using Monte-Carlo methods
http://arxiv.org/abs/2006.04176
AUTHORS:        Zafeirios Fountas ; Noor Sajid ; Pedro A. M. Mediano ; Karl Friston
COMMENTS:       16 pages, 8 figures
HIGHLIGHT:      In this paper, we present a neural architecture for building deep active inference agents operating in complex, continuous state-spaces using multiple forms of Monte-Carlo (MC) sampling.

6, TITLE:       An Empirical Meta-analysis of the Life Sciences (Linked?) Open Data on the Web
http://arxiv.org/abs/2006.04161
AUTHORS:        Maulik R. Kamdar ; Mark A. Musen
COMMENTS:       Under Review at Nature Scientific Data
HIGHLIGHT:      In this paper, we extract schemas from more than 80 publicly available biomedical linked data graphs into an LSLOD schema graph and conduct an empirical meta-analysis to evaluate the extent of semantic heterogeneity across the LSLOD cloud.

7, TITLE:       A data complexity and rewritability tetrachotomy of ontology-mediated queries with a covering axiom
http://arxiv.org/abs/2006.04167
AUTHORS:        Olga Gerasimova ; Stanislav Kikot ; Agi Kurucz ; Vladimir Podolskii ; Michael Zakharyaschev
HIGHLIGHT:      Aiming to understand the data complexity of answering conjunctive queries mediated by an axiom stating that a class is covered by the union of two other classes, we show that deciding their first-order rewritability is PSpace-hard and obtain a number of sufficient conditions for membership in AC0, L, NL, and P.

8, TITLE:       Unsupervised Abnormality Detection Using Heterogeneous Autonomous Systems
http://arxiv.org/abs/2006.03733
AUTHORS:        Sayeed Shafayet Chowdhury ; Kazi Mejbaul Islam ; Rouhan Noor
HIGHLIGHT:      In this paper, a heterogeneous system is proposed which estimates the degree of abnormality of an environment using drone-feed, analyzing real-time image and IMU sensor data in an unsupervised manner.

9, TITLE:       WOAD: Weakly Supervised Online Action Detection in Untrimmed Videos
http://arxiv.org/abs/2006.03732
AUTHORS:        Mingfei Gao ; Yingbo Zhou ; Ran Xu ; Richard Socher ; Caiming Xiong
HIGHLIGHT:      We propose WOAD, a weakly supervised framework that can be trained using only video-class labels.

10, TITLE:      Applied Awareness: Test-Driven GUI Development using Computer Vision and Cryptography
http://arxiv.org/abs/2006.03725
AUTHORS:        Donald Beaver
COMMENTS:       8 pages. Submitted for conference publication
HIGHLIGHT:      Our interdisciplinary work is ready for off-the-shelf practice: we report self-contained, practical implementation with both online and offline validation, using simple designer specifications at the outset and specifically avoiding any requirements for a bootstrap implementation or golden images.

11, TITLE:      State Action Separable Reinforcement Learning
http://arxiv.org/abs/2006.03713
AUTHORS:        Ziyao Zhang ; Liang Ma ; Kin K. Leung ; Konstantinos Poularakis ; Mudhakar Srivatsa
COMMENTS:       16 pages
HIGHLIGHT:      In this regard, we propose a new learning paradigm, State Action Separable Reinforcement Learning (sasRL), wherein the action space is decoupled from the value function learning process for higher efficiency.

12, TITLE:      Relation of the Relations: A New Paradigm of the Relation Extraction Problem
http://arxiv.org/abs/2006.03719
AUTHORS:        Zhijing Jin ; Yongyi Yang ; Xipeng Qiu ; Zheng Zhang
HIGHLIGHT:      Due to the significance of RoR in existing datasets, we propose a new paradigm of RE that considers as a whole the predictions of all relations in the same context.

13, TITLE:      Accelerating Natural Language Understanding in Task-Oriented Dialog
http://arxiv.org/abs/2006.03701
AUTHORS:        Ojas Ahuja ; Shrey Desai
COMMENTS:       Accepted to ACL 2020 Workshop on NLP for Conversational AI
HIGHLIGHT:      In this work, we show that a simple convolutional model compressed with structured pruning achieves largely comparable results to BERT on ATIS and Snips, with under 100K parameters.

14, 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.

15, TITLE:      FibeR-CNN: Expanding Mask R-CNN to Improve Image-Based Fiber Analysis
http://arxiv.org/abs/2006.04552
AUTHORS:        Max Frei ; Frank Einar Kruis
COMMENTS:       21 pages, 30 figures, 5 tables, 1 algorithm
HIGHLIGHT:      We therefore propose the use of region-based convolutional neural networks (R-CNNs) to automate this task.

16, TITLE:      Texture Interpolation for Probing Visual Perception
http://arxiv.org/abs/2006.03698
AUTHORS:        Jonathan Vacher ; Aida Davila ; Adam Kohn ; Ruben Coen-Cagli
COMMENTS:       Paper + ref: 12 pages and 6 figures | Supplementary: 8 pages and 8 figures
HIGHLIGHT:      We demonstrate our method by measuring the perceptual scale associated to the interpolation parameter in human observers, and the neural sensitivity of different areas of visual cortex in macaque monkeys.

17, TITLE:      UDPipe at EvaLatin 2020: Contextualized Embeddings and Treebank Embeddings
http://arxiv.org/abs/2006.03687
AUTHORS:        Milan Straka ; Jana Straková
COMMENTS:       Accepted at EvaLatin 2020, LREC (Proceedings of Language Resources and Evaluation, Marseille, France)
HIGHLIGHT:      We present our contribution to the EvaLatin shared task, which is the first evaluation campaign devoted to the evaluation of NLP tools for Latin.

18, TITLE:      Evaluating the Disentanglement of Deep Generative Models through Manifold Topology
http://arxiv.org/abs/2006.03680
AUTHORS:        Sharon Zhou ; Eric Zelikman ; Fred Lu ; Andrew Y. Ng ; Stefano Ermon
HIGHLIGHT:      To address this, we present a method for quantifying disentanglement that only uses the generative model, by measuring the topological similarity of conditional submanifolds in the learned representation.

19, TITLE:      Prague Dependency Treebank -- Consolidated 1.0
http://arxiv.org/abs/2006.03679
AUTHORS:        Jan Hajič ; Eduard Bejček ; Jaroslava Hlaváčová ; Marie Mikulová ; Milan Straka ; Jan Štěpánek ; Barbora Štěpánková
COMMENTS:       Accepted at LREC 2020 (Proceedings of Language Resources and Evaluation, Marseille, France)
HIGHLIGHT:      We present a richly annotated and genre-diversified language resource, the Prague Dependency Treebank-Consolidated 1.0 (PDT-C 1.0), the purpose of which is - as it always been the case for the family of the Prague Dependency Treebanks - to serve both as a training data for various types of NLP tasks as well as for linguistically-oriented research.

20, 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.

21, 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.

22, TITLE:      More Information Supervised Probabilistic Deep Face Embedding Learning
http://arxiv.org/abs/2006.04518
AUTHORS:        Ying Huang ; Shangfeng Qiu ; Wenwei Zhang ; Xianghui Luo ; Jinzhuo Wang
HIGHLIGHT:      In this paper, we analyse margin based softmax loss in probability view.

23, TITLE:      Combining word embeddings and convolutional neural networks to detect duplicated questions
http://arxiv.org/abs/2006.04513
AUTHORS:        Yoan Dimitrov
HIGHLIGHT:      In this work, we propose a simple approach to identifying semantically similar questions by combining the strengths of word embeddings and Convolutional Neural Networks (CNNs).

24, TITLE:      VectorTSP: A Traveling Salesperson Problem with Racetrack-like acceleration constraints
http://arxiv.org/abs/2006.03666
AUTHORS:        Arnaud Casteigts ; Mathieu Raffinot ; Jason Schoeters
COMMENTS:       12 pages, 22 pages with bibliography and appendices, submitted
HIGHLIGHT:      We present algorithms for both, and we demonstrate and quantify through experiments that this approach frequently finds a better solution than the optimal trajectory realizing an optimal ETSP tour, which legitimates the problem itself and (we hope) motivates further algorithmic developments.

25, TITLE:      Rapid Task-Solving in Novel Environments
http://arxiv.org/abs/2006.03662
AUTHORS:        Sam Ritter ; Ryan Faulkner ; Laurent Sartran ; Adam Santoro ; Matt Botvinick ; David Raposo
HIGHLIGHT:      We introduce two domains for conducting research on this challenge, and find that state-of-the-art deep reinforcement learning (RL) agents fail to plan in novel environments.

26, 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.

27, TITLE:      DeBERTa: Decoding-enhanced BERT with Disentangled Attention
http://arxiv.org/abs/2006.03654
AUTHORS:        Pengcheng He ; Xiaodong Liu ; Jianfeng Gao ; Weizhu Chen
COMMENTS:       17 pages,4 figures, 8 tables
HIGHLIGHT:      In this paper we propose a new model architecture DeBERTa (Decoding-enhanced BERT with disentangled attention) that improves the BERT and RoBERTa models using two novel techniques.

28, TITLE:      AutoHAS: Differentiable Hyper-parameter and Architecture Search
http://arxiv.org/abs/2006.03656
AUTHORS:        Xuanyi Dong ; Mingxing Tan ; Adams Wei Yu ; Daiyi Peng ; Bogdan Gabrys ; Quoc V. Le
HIGHLIGHT:      In this work, we extend NAS to a broader and more practical space by combining hyper-parameter and architecture search.

29, TITLE:      Filtered Inner Product Projection for Multilingual Embedding Alignment
http://arxiv.org/abs/2006.03652
AUTHORS:        Vin Sachidananda ; Ziyi Yang ; Chenguang Zhu
HIGHLIGHT:      In this paper, we propose a method, Filtered Inner Product Projection (FIPP), for mapping embeddings to a common representation space and evaluate FIPP in the context of bilingual dictionary induction.

30, TITLE:      Deployment-Efficient Reinforcement Learning via Model-Based Offline Optimization
http://arxiv.org/abs/2006.03647
AUTHORS:        Tatsuya Matsushima ; Hiroki Furuta ; Yutaka Matsuo ; Ofir Nachum ; Shixiang Gu
HIGHLIGHT:      We propose a novel model-based algorithm, Behavior-Regularized Model-ENsemble (BREMEN) that can effectively optimize a policy offline using 10-20 times fewer data than prior works.

31, TITLE:      RIT-Eyes: Rendering of near-eye images for eye-tracking applications
http://arxiv.org/abs/2006.03642
AUTHORS:        Nitinraj Nair ; Rakshit Kothari ; Aayush K. Chaudhary ; Zhizhuo Yang ; Gabriel J. Diaz ; Jeff B. Pelz ; Reynold J. Bailey
HIGHLIGHT:      In this work, we introduce a synthetic eye image generation platform that improves upon previous work by adding features such as an active deformable iris, an aspherical cornea, retinal retro-reflection, gaze-coordinated eye-lid deformations, and blinks.

32, TITLE:      Stance Detection on Social Media: State of the Art and Trends
http://arxiv.org/abs/2006.03644
AUTHORS:        Abeer AlDayel ; Walid Magdy
HIGHLIGHT:      The survey reports the state-of-the-art results on the existing benchmark datasets onstance detection, and discusses the most effective approaches.In addition, this study explores the emerging trends and the different applications of stance detection on social media.The study concludes by providing discussion of the gabs in the current existing research and highlighting the possible futuredirections for stance detection on social media

33, TITLE:      Knowledge transfer between bridges for drive-by monitoring using adversarial and multi-task learning
http://arxiv.org/abs/2006.03641
AUTHORS:        Jingxiao Liu ; Mario Bergés ; Jacobo Bielak ; Hae Young Noh
HIGHLIGHT:      To alleviate these problems, we introduce a transfer learning framework using domain-adversarial training and multi-task learning to detect, localize and quantify damage.

34, TITLE:      Robust Face Verification via Disentangled Representations
http://arxiv.org/abs/2006.03638
AUTHORS:        Marius Arvinte ; Ahmed H. Tewfik ; Sriram Vishwanath
COMMENTS:       Preprint
HIGHLIGHT:      We introduce a robust algorithm for face verification, i.e., deciding whether twoimages are of the same person or not.

35, TITLE:      SparseFusion: Dynamic Human Avatar Modeling from Sparse RGBD Images
http://arxiv.org/abs/2006.03630
AUTHORS:        Xinxin Zuo ; Sen Wang ; Jiangbin Zheng ; Weiwei Yu ; Minglun Gong ; Ruigang Yang ; Li Cheng
COMMENTS:       Accepted by TMM
HIGHLIGHT:      In this paper, we propose a novel approach to reconstruct 3D human body shapes based on a sparse set of RGBD frames using a single RGBD camera.

36, TITLE:      Equivariant Maps for Hierarchical Structures
http://arxiv.org/abs/2006.03627
AUTHORS:        Renhao Wang ; Marjan Albooyeh ; Siamak Ravanbakhsh
HIGHLIGHT:      Observing that the transformation group for a nested structure corresponds to the ``wreath product'' of the symmetry groups of the building blocks, we show how to obtain the equivariant map for hierarchical data-structures using an intuitive combination of the equivariant maps for the individual blocks.

37, TITLE:      LGML: Logic Guided Machine Learning
http://arxiv.org/abs/2006.03626
AUTHORS:        Joseph Scott ; Maysum Panju ; Vijay Ganesh
HIGHLIGHT:      We introduce Logic Guided Machine Learning (LGML), a novel approach that symbiotically combines machine learning (ML) and logic solvers with the goal of learning mathematical functions from data.

38, TITLE:      Inception Augmentation Generative Adversarial Network
http://arxiv.org/abs/2006.03622
AUTHORS:        Saman Motamed ; Farzad Khalvati
HIGHLIGHT:      In this work, we propose a new GAN architecture for semi-supervised augmentation of chest X-rays for the detection of pneumonia.

39, TITLE:      Hierarchical Class-Based Curriculum Loss
http://arxiv.org/abs/2006.03629
AUTHORS:        Palash Goyal ; Shalini Ghosh
HIGHLIGHT:      In this paper, we propose a loss function, hierarchical curriculum loss, with two properties: (i) satisfy hierarchical constraints present in the label space, and (ii) provide non-uniform weights to labels based on their levels in the hierarchy, learned implicitly by the training paradigm.

40, TITLE:      Decentralised Learning from Independent Multi-Domain Labels for Person Re-Identification
http://arxiv.org/abs/2006.04150
AUTHORS:        Guile Wu ; Shaogang Gong
HIGHLIGHT:      In this work, we solve the Re-ID problem by decentralised model learning from non-shared private training data distributed at multiple user cites of independent multi-domain labels.

41, TITLE:      Analogy as Nonparametric Bayesian Inference over Relational Systems
http://arxiv.org/abs/2006.04156
AUTHORS:        Ruairidh M. Battleday ; Thomas L. Griffiths
COMMENTS:       In Proceedings for the Annual Meeting of the Cognitive Science Society 2020 (CogSci 2020)
HIGHLIGHT:      In this project, we propose a Bayesian model that generalizes relational knowledge to novel environments by analogically weighting predictions from previously encountered relational structures.

42, TITLE:      BERT Loses Patience: Fast and Robust Inference with Early Exit
http://arxiv.org/abs/2006.04152
AUTHORS:        Wangchunshu Zhou ; Canwen Xu ; Tao Ge ; Julian McAuley ; Ke Xu ; Furu Wei
COMMENTS:       13 pages
HIGHLIGHT:      In this paper, we propose Patience-based Early Exit, a straightforward yet effective inference method that can be used as a plug-and-play technique to simultaneously improve the efficiency and robustness of a pretrained language model (PLM).

43, TITLE:      Interactive Extractive Search over Biomedical Corpora
http://arxiv.org/abs/2006.04148
AUTHORS:        Hillel Taub-Tabib ; Micah Shlain ; Shoval Sadde ; Dan Lahav ; Matan Eyal ; Yaara Cohen ; Yoav Goldberg
HIGHLIGHT:      We present a system that allows life-science researchers to search a linguistically annotated corpus of scientific texts using patterns over dependency graphs, as well as using patterns over token sequences and a powerful variant of boolean keyword queries.

44, TITLE:      Peer Collaborative Learning for Online Knowledge Distillation
http://arxiv.org/abs/2006.04147
AUTHORS:        Guile Wu ; Shaogang Gong
HIGHLIGHT:      In this work, we propose a novel Peer Collaborative Learning method for online knowledge distillation.

45, TITLE:      Parametric Representation for Singing Voice Synthesis: a Comparative Evaluation
http://arxiv.org/abs/2006.04142
AUTHORS:        Onur Babacan ; Thomas Drugman ; Tuomo Raitio ; Daniel Erro ; Thierry Dutoit
HIGHLIGHT:      The goal of this paper is twofold.

46, TITLE:      Maximum Phase Modeling for Sparse Linear Prediction of Speech
http://arxiv.org/abs/2006.04138
AUTHORS:        Thomas Drugman
HIGHLIGHT:      The aim of this paper is to propose a novel technique which incorporates a modeling of the maximum-phase contribution of speech, and can be applied to any filter representation.

47, TITLE:      Learning Texture Transformer Network for Image Super-Resolution
http://arxiv.org/abs/2006.04139
AUTHORS:        Fuzhi Yang ; Huan Yang ; Jianlong Fu ; Hongtao Lu ; Baining Guo
COMMENTS:       Accepted by CVPR 2020
HIGHLIGHT:      In this paper, we propose a novel Texture Transformer Network for Image Super-Resolution (TTSR), in which the LR and Ref images are formulated as queries and keys in a transformer, respectively.

48, TITLE:      Analysis and Synthesis of Hypo and Hyperarticulated Speech
http://arxiv.org/abs/2006.04136
AUTHORS:        Benjamin Picart ; Thomas Drugman ; Thierry Dutoit
HIGHLIGHT:      This paper focuses on the analysis and synthesis of hypo and hyperarticulated speech in the framework of HMM-based speech synthesis.

49, TITLE:      ADMP: An Adversarial Double Masks Based Pruning Framework For Unsupervised Cross-Domain Compression
http://arxiv.org/abs/2006.04127
AUTHORS:        Xiaoyu Feng ; Zhuqing Yuan ; Guijin Wang ; Yongpan Liu
HIGHLIGHT:      Hence this work proposes an Adversarial Double Masks based Pruning (ADMP) for such cross-domain compression.

50, TITLE:      A multi-channel framework for joint reconstruction of multi-contrast parallel MRI
http://arxiv.org/abs/2006.04128
AUTHORS:        Erfan Ebrahim Esfahani
HIGHLIGHT:      In this paper, a novel isotropic multi-channel image regularizer is introduced, and its full potential is unleashed by integrating it into compressed multi-contrast multi-coil MRI.

51, TITLE:      On the Complexity of Branching Proofs
http://arxiv.org/abs/2006.04124
AUTHORS:        Daniel Dadush ; Samarth Tiwari
HIGHLIGHT:      As our final contribution, we give a simple family of polytopes in $[0,1]^n$ requiring branching proofs of length $2^n/n$.

52, TITLE:      Sophisticated Inference
http://arxiv.org/abs/2006.04120
AUTHORS:        Karl Friston ; Lancelot Da Costa ; Danijar Hafner ; Casper Hesp ; Thomas Parr
HIGHLIGHT:      In this paper, we consider a sophisticated kind of active inference, using a recursive form of expected free energy.

53, TITLE:      DiffGCN: Graph Convolutional Networks via Differential Operators and Algebraic Multigrid Pooling
http://arxiv.org/abs/2006.04115
AUTHORS:        Moshe Eliasof ; Eran Treister
HIGHLIGHT:      In this work we propose novel approaches for graph convolution, pooling and unpooling, taking inspiration from finite-elements and algebraic multigrid frameworks.

54, TITLE:      Iterative Effect-Size Bias in Ridehailing: Measuring Social Bias in Dynamic Pricing of 100 Million Rides
http://arxiv.org/abs/2006.04599
AUTHORS:        Akshat Pandey ; Aylin Caliskan
COMMENTS:       10 pages, 3 tables, 2 figures
HIGHLIGHT:      In this work we develop a random-effects based metric for the analysis of social bias in supervised machine learning prediction models where model outputs depend on U.S. locations.

55, TITLE:      WaveNODE: A Continuous Normalizing Flow for Speech Synthesis
http://arxiv.org/abs/2006.04598
AUTHORS:        Hyeongju Kim ; Hyeongseung Lee ; Woo Hyun Kang ; Sung Jun Cheon ; Byoung Jin Choi ; Nam Soo Kim
COMMENTS:       8 pages, 4 figures
HIGHLIGHT:      In this paper, we propose a novel generative model called WaveNODE which exploits a continuous normalizing flow for speech synthesis.

56, TITLE:      CS-Embed-francesita at SemEval-2020 Task 9: The effectiveness of code-switched word embeddings for sentiment analysis
http://arxiv.org/abs/2006.04597
AUTHORS:        Frances Adriana Laureano De Leon ; Florimond Guéniat ; Harish Tayyar Madabushi
COMMENTS:       To appear in SemEval-2020 Task 9
HIGHLIGHT:      In this work, we present word-embeddings trained on code-switched tweets, specifically those that make use of Spanish and English, known as Spanglish.

57, TITLE:      Incorporating Pragmatic Reasoning Communication into Emergent Language
http://arxiv.org/abs/2006.04109
AUTHORS:        Yipeng Kang ; Tonghan Wang ; Gerard de Melo
COMMENTS:       8 pages. Submitted to NeurIPS 2020
HIGHLIGHT:      Given that their combination has been explored in linguistics, we propose computational models that combine short-term mutual reasoning-based pragmatics with long-term language emergentism.

58, TITLE:      Language Models as Fact Checkers?
http://arxiv.org/abs/2006.04102
AUTHORS:        Nayeon Lee ; Belinda Z. Li ; Sinong Wang ; Wen-tau Yih ; Hao Ma ; Madian Khabsa
COMMENTS:       Accepted in FEVER Workshop (ACL2020)
HIGHLIGHT:      In this paper, we leverage this implicit knowledge to create an effective end-to-end fact checker using a solely a language model, without any external knowledge or explicit retrieval components.

59, TITLE:      Person Re-identification in the 3D Space
http://arxiv.org/abs/2006.04569
AUTHORS:        Zhedong Zheng ; Yi Yang
COMMENTS:       The code is available at https://github.com/layumi/person-reid-3d
HIGHLIGHT:      In this work, we address this limitation by exploring the prior knowledge of the 3D body structure.

60, TITLE:      Towards an Argument Mining Pipeline Transforming Texts to Argument Graphs
http://arxiv.org/abs/2006.04562
AUTHORS:        Mirko Lenz ; Premtim Sahitaj ; Sean Kallenberg ; Christopher Coors ; Lorik Dumani ; Ralf Schenkel ; Ralph Bergmann
HIGHLIGHT:      We present an argument mining pipeline as a universally applicable approach for transforming German and English language texts to graph-based argument representations.

61, TITLE:      FastSpeech 2: Fast and High-Quality End-to-End Text-to-Speech
http://arxiv.org/abs/2006.04558
AUTHORS:        Yi Ren ; Chenxu Hu ; Tao Qin ; Sheng Zhao ; Zhou Zhao ; Tie-Yan Liu
HIGHLIGHT:      In this paper, we propose FastSpeech 2, which addresses the issues in FastSpeech and better solves the one-to-many mapping problem in TTS by 1) directly training the model with ground-truth target instead of the simplified output from teacher, and 2) introducing more variation information of speech (e.g., pitch, energy and more accurate duration) as conditional inputs.

62, TITLE:      Multi-view Contrastive Learning for Online Knowledge Distillation
http://arxiv.org/abs/2006.04093
AUTHORS:        Chuanguang Yang ; Zhulin An ; Xiaolong Hu ; Hui Zhu ; Kaiqiang Xu ; Yongjun Xu
HIGHLIGHT:      We thus introduce Multi-view Contrastive Learning (MCL) for OKD to implicitly capture correlations of representations encoded by multiple peer networks, which provide various views for understanding the input data samples.

63, TITLE:      Robust Learning Through Cross-Task Consistency
http://arxiv.org/abs/2006.04096
AUTHORS:        Amir Zamir ; Alexander Sax ; Teresa Yeo ; Oğuzhan Kar ; Nikhil Cheerla ; Rohan Suri ; Zhangjie Cao ; Jitendra Malik ; Leonidas Guibas
COMMENTS:       CVPR 2020 (Oral). Project website, models, live demo at http://consistency.epfl.ch/
HIGHLIGHT:      We propose a broadly applicable and fully computational method for augmenting learning with Cross-Task Consistency.

64, TITLE:      End-to-end Learning for Inter-Vehicle Distance and Relative Velocity Estimation in ADAS with a Monocular Camera
http://arxiv.org/abs/2006.04082
AUTHORS:        Zhenbo Song ; Jianfeng Lu ; Tong Zhang ; Hongdong Li
HIGHLIGHT:      In this paper, we propose a monocular camera based inter-vehicle distance and relative velocity estimation method based on end-to-end training of a deep neural network.

65, TITLE:      CubifAE-3D: Monocular Camera Space Cubification on Autonomous Vehicles for Auto-Encoder based 3D Object Detection
http://arxiv.org/abs/2006.04080
AUTHORS:        Shubham Shrivastava ; Punarjay Chakravarty
COMMENTS:       11 pages, 14 figures
HIGHLIGHT:      We introduce a method for 3D object detection using a single monocular image.

66, TITLE:      STDI-Net: Spatial-Temporal Network with Dynamic Interval Mapping for Bike Sharing Demand Prediction
http://arxiv.org/abs/2006.04089
AUTHORS:        Weiguo Pian ; Yingbo Wu ; Ziyi Kou
COMMENTS:       submitted to ACML2020
HIGHLIGHT:      In this paper, we propose a novel deep learning method called Spatial-Temporal Dynamic Interval Network (STDI-Net).

67, TITLE:      Implications of Human Irrationality for Reinforcement Learning
http://arxiv.org/abs/2006.04072
AUTHORS:        Haiyang Chen ; Hyung Jin Chang ; Andrew Howes
COMMENTS:       12 pages, 5 figures
HIGHLIGHT:      Here, we propose a novel POMDP model for contextual choice tasks and show that, despite the apparent irrationalities, a reinforcement learner can take advantage of the way that humans make decisions.

68, TITLE:      Siamese Keypoint Prediction Network for Visual Object Tracking
http://arxiv.org/abs/2006.04078
AUTHORS:        Qiang Li ; Zekui Qin ; Wenbo Zhang ; Wen Zheng
COMMENTS:       Code: https://github.com/ZekuiQin/SiamKPN
HIGHLIGHT:      In this paper, we propose the Siamese keypoint prediction network (SiamKPN) to address these challenges.

69, TITLE:      Dual Policy Distillation
http://arxiv.org/abs/2006.04061
AUTHORS:        Kwei-Herng Lai ; Daochen Zha ; Yuening Li ; Xia Hu
COMMENTS:       Accepeted by IJCAI'20
HIGHLIGHT:      In this work, we introduce dual policy distillation(DPD), a student-student framework in which two learners operate on the same environment to explore different perspectives of the environment and extract knowledge from each other to enhance their learning.

70, TITLE:      Growing Together: Modeling Human Language Learning With n-Best Multi-Checkpoint Machine Translation
http://arxiv.org/abs/2006.04050
AUTHORS:        El Moatez Billah Nagoudi ; Muhammad Abdul-Mageed ; Hasan Cavusoglu
COMMENTS:       Accepted to the 4th Workshop on Neural Generation and Translation (Duolingo Shared Task on Simultaneous Translation And Paraphrase for Language Education Mayhew et al., 2020) collocated with ACL 2020
HIGHLIGHT:      We describe our submission to the 2020 Duolingo Shared Task on Simultaneous Translation And Paraphrase for Language Education (STAPLE) (Mayhew et al., 2020).

71, TITLE:      Facial Expression Recognition using Deep Learning
http://arxiv.org/abs/2006.04057
AUTHORS:        Raghu Vamshi. N ; Bharathi Raja S
HIGHLIGHT:      In this paper, I take one such dataset FER-2013 and will implement deep learning models that are able to achieve significant improvement over the previously used traditional approaches and even some of the deep learning models.

72, TITLE:      NITS-VC System for VATEX Video Captioning Challenge 2020
http://arxiv.org/abs/2006.04058
AUTHORS:        Alok Singh ; Thoudam Doren Singh ; Sivaji Bandyopadhyay
COMMENTS:       4 pages, 1 figure
HIGHLIGHT:      In this paper, a system description of the framework used for VATEX-2020 video captioning challenge is presented.

73, TITLE:      A Generic First-Order Algorithmic Framework for Bi-Level Programming Beyond Lower-Level Singleton
http://arxiv.org/abs/2006.04045
AUTHORS:        Risheng Liu ; Pan Mu ; Xiaoming Yuan ; Shangzhi Zeng ; Jin Zhang
COMMENTS:       Accepted at ICML 2020
HIGHLIGHT:      Theoretically, we derive a new methodology to prove the convergence of BDA without the LLS condition.

74, TITLE:      DeepRelativeFusion: Dense Monocular SLAM using Single-Image Relative Depth Prediction
http://arxiv.org/abs/2006.04047
AUTHORS:        Shing Yan Loo ; Syamsiah Mashohor ; Sai Hong Tang ; Hong Zhang
HIGHLIGHT:      In this paper, we propose a dense monocular SLAM system, named DeepRelativeFusion, that is capable to recover a globally consistent 3D structure.

75, TITLE:      An Algorithm for Fuzzification of WordNets, Supported by a Mathematical Proof
http://arxiv.org/abs/2006.04042
AUTHORS:        Sayyed-Ali Hossayni ; Mohammad-R Akbarzadeh-T ; Diego Reforgiato Recupero ; Aldo Gangemi ; Esteve Del Acebo ; Josep Lluís de la Rosa i Esteva
COMMENTS:       6 pages, without figures, theoretical
HIGHLIGHT:      In this study, we present an algorithm for constructing fuzzy versions of WLDs of any language, given a corpus of documents and a word-sense disambiguation (WSD) system for that language.

76, TITLE:      SVGA-Net: Sparse Voxel-Graph Attention Network for 3D Object Detection from Point Clouds
http://arxiv.org/abs/2006.04043
AUTHORS:        Qingdong He ; Zhengning Wang ; Hao Zeng ; Yi Zeng ; Shuaicheng Liu ; Bing Zeng
HIGHLIGHT:      In this paper, we propose Sparse Voxel-Graph Attention Network (SVGA-Net), a novel end-to-end trainable network which mainly contains voxel-graph module and sparse-to-dense regression module to achieve comparable 3D detection tasks from raw LIDAR data.

77, TITLE:      Metagame Autobalancing for Competitive Multiplayer Games
http://arxiv.org/abs/2006.04419
AUTHORS:        Daniel Hernandez ; Charles Takashi Toyin Gbadomosi ; James Goodman ; James Alfred Walker
HIGHLIGHT:      In this paper we present a tool for balancing multi-player games during game design.

78, TITLE:      A Survey on Approximation in Parameterized Complexity: Hardness and Algorithms
http://arxiv.org/abs/2006.04411
AUTHORS:        Andreas Emil Feldmann ; Karthik C. S. ; Euiwoong Lee ; Pasin Manurangsi
HIGHLIGHT:      We survey developments in the area both from the algorithmic and hardness perspectives, with emphasis on new techniques and potential future research directions.

79, TITLE:      Passive Batch Injection Training Technique: Boosting Network Performance by Injecting Mini-Batches from a different Data Distribution
http://arxiv.org/abs/2006.04406
AUTHORS:        Pravendra Singh ; Pratik Mazumder ; Vinay P. Namboodiri
COMMENTS:       Accepted in IJCNN 2020
HIGHLIGHT:      This work presents a novel training technique for deep neural networks that makes use of additional data from a distribution that is different from that of the original input data.

80, TITLE:      An Optimal Tester for $k$-Linear
http://arxiv.org/abs/2006.04409
AUTHORS:        Nader H. Bshouty
HIGHLIGHT:      In this paper, we study property testing of the classes $k$-Linear, the class of all $k$-linear functions, and $k$-Linear$^*$, the class $\cup_{j=0}^kj$-Linear.

81, TITLE:      MeshSDF: Differentiable Iso-Surface Extraction
http://arxiv.org/abs/2006.03997
AUTHORS:        Edoardo Remelli ; Artem Lukoianov ; Stephan R. Richter ; Benoît Guillard ; Timur Bagautdinov ; Pierre Baque ; Pascal Fua
COMMENTS:       11 pages, 5 figures, submitted to NeurIPS 2020
HIGHLIGHT:      In this work, we remove this limitation and introduce a differentiable way to produce explicit surface mesh representations from Deep Signed Distance Functions.

82, TITLE:      Graph Neural Network Encoding for Community Detection in Attribute Networks
http://arxiv.org/abs/2006.03996
AUTHORS:        Jianyong Sun ; Wei Zheng ; Qingfu Zhang ; Zongben Xu
HIGHLIGHT:      The fitness landscape analysis verifies that the transformed community detection problems have smoother landscapes than those of the original problems, which justifies the effectiveness of the proposed graph neural network encoding method.

83, TITLE:      Enhancing Facial Data Diversity with Style-based Face Aging
http://arxiv.org/abs/2006.03985
AUTHORS:        Markos Georgopoulos ; James Oldfield ; Mihalis A. Nicolaou ; Yannis Panagakis ; Maja Pantic
COMMENTS:       IEEE CVPR 2020 WORKSHOP ON FAIR, DATA EFFICIENT AND TRUSTED COMPUTER VISION
HIGHLIGHT:      In this work, we address the problem of increasing the diversity of face datasets with respect to age.

84, TITLE:      Reinforcement Learning for Multi-Product Multi-Node Inventory Management in Supply Chains
http://arxiv.org/abs/2006.04037
AUTHORS:        Nazneen N Sultana ; Hardik Meisheri ; Vinita Baniwal ; Somjit Nath ; Balaraman Ravindran ; Harshad Khadilkar
HIGHLIGHT:      We describe a novel formulation in a multi-agent (hierarchical) reinforcement learning framework that can be used for parallelised decision-making, and use the advantage actor critic (A2C) algorithm with quantised action spaces to solve the problem.

85, TITLE:      Semantic Loss Application to Entity Relation Recognition
http://arxiv.org/abs/2006.04031
AUTHORS:        Venkata Sasank Pagolu
HIGHLIGHT:      The main contribution of this paper is an end-to-end neural model for joint entity relation extraction which incorporates a novel loss function.

86, TITLE:      A Comparative Analysis of E-Scooter and E-Bike Usage Patterns: Findings from the City of Austin, TX
http://arxiv.org/abs/2006.04033
AUTHORS:        Mohammed Hamad Almannaa ; Huthaifa I. Ashqar ; Mohammed Elhenawy ; Mahmoud Masoud ; Andry Rakotonirainy ; Hesham Rakha
COMMENTS:       Submitted to the International Journal of Sustainable Transportation
HIGHLIGHT:      This study analyzes e-scooter and dockless e-bike sharing system user behavior.

87, TITLE:      Efficient Architecture Search for Continual Learning
http://arxiv.org/abs/2006.04027
AUTHORS:        Qiang Gao ; Zhipeng Luo ; Diego Klabjan
COMMENTS:       12 pages, 11 figures
HIGHLIGHT:      To reach these goals, we propose a novel approach named as Continual Learning with Efficient Architecture Search, or CLEAS in short.

88, TITLE:      SharinGAN: Combining Synthetic and Real Data for Unsupervised Geometry Estimation
http://arxiv.org/abs/2006.04026
AUTHORS:        Koutilya PNVR ; Hao Zhou ; David Jacobs
COMMENTS:       Accepted to CVPR 2020. Supplementary material added towards the end instead of a separate file. A Github link to the code is also provided in this submission
HIGHLIGHT:      We propose a novel method for combining synthetic and real images when training networks to determine geometric information from a single image.

89, TITLE:      A Multitask Learning Approach for Diacritic Restoration
http://arxiv.org/abs/2006.04016
AUTHORS:        Sawsan Alqahtani ; Ajay Mishra ; Mona Diab
HIGHLIGHT:      Thus, to compensate for this loss, we investigate the use of multi-task learning to jointly optimize diacritic restoration with related NLP problems namely word segmentation, part-of-speech tagging, and syntactic diacritization.

90, TITLE:      Learning under Invariable Bayesian Safety
http://arxiv.org/abs/2006.04497
AUTHORS:        Gal Bahar ; Omer Ben-Porat ; Kevin Leyton-Brown ; Moshe Tennenholtz
HIGHLIGHT:      In this paper, we adopt a model inspired by recent work on a bandit-like setting for recommendations.

91, TITLE:      Medical Concept Normalization in User Generated Texts by Learning Target Concept Embeddings
http://arxiv.org/abs/2006.04014
AUTHORS:        Katikapalli Subramanyam Kalyan ; S. Sangeetha
COMMENTS:       5 pages
HIGHLIGHT:      Our proposed model overcomes these drawbacks by jointly learning the representations of input concept mention and target concepts.

92, TITLE:      Action Recognition with Deep Multiple Aggregation Networks
http://arxiv.org/abs/2006.04489
AUTHORS:        Ahmed Mazari ; Hichem Sahbi
HIGHLIGHT:      In this paper, we introduce a novel hierarchical pooling design that captures different levels of temporal granularity in action recognition.

93, TITLE:      Neural Networks Out-of-Distribution Detection: Hyperparameter-Free Isotropic Maximization Loss, The Principle of Maximum Entropy, Cold Training, and Branched Inferences
http://arxiv.org/abs/2006.04005
AUTHORS:        David Macêdo ; Teresa Ludermir
HIGHLIGHT:      In this paper, we propose a novel loss called Hyperparameter-Free IsoMax that overcomes these limitations.

94, TITLE:      Real-Time Model Calibration with Deep Reinforcement Learning
http://arxiv.org/abs/2006.04001
AUTHORS:        Manuel Arias Chao ; Yuan Tian ; Chetan Kulkarni ; Kai Goebel ; Olga Fink
COMMENTS:       18 pages, 10 figures
HIGHLIGHT:      In this paper, we propose a novel framework for inference of model parameters based on reinforcement learning.

95, TITLE:      Every Action Based Sensor
http://arxiv.org/abs/2006.04003
AUTHORS:        Grace McFassel ; Dylan A. Shell
COMMENTS:       16 pages, 7 figures, WAFR 2020
HIGHLIGHT:      Erdmann's theory of action-based sensors is a classical approach to characterizing fundamental information requirements.

96, TITLE:      A Comparison of Self-Play Algorithms Under a Generalized Framework
http://arxiv.org/abs/2006.04471
AUTHORS:        Daniel Hernandez ; Kevin Denamganai ; Sam Devlin ; Spyridon Samothrakis ; James Alfred Walker
HIGHLIGHT:      We present a formalized framework, with clearly defined assumptions, which encapsulates the meaning of self-play as abstracted from various existing self-play algorithms.

97, TITLE:      Deep hierarchical pooling design for cross-granularity action recognition
http://arxiv.org/abs/2006.04473
AUTHORS:        Ahmed Mazari ; Hichem Sahbi
HIGHLIGHT:      In this paper, we introduce a novel hierarchical aggregation design that captures different levels of temporal granularity in action recognition.

98, TITLE:      A non-causal FFTNet architecture for speech enhancement
http://arxiv.org/abs/2006.04469
AUTHORS:        Muhammed PV Shifas ; Nagaraj Adiga ; Vassilis Tsiaras ; Yannis Stylianou
COMMENTS:       5 pages
HIGHLIGHT:      In this paper, we suggest a new parallel, non-causal and shallow waveform domain architecture for speech enhancement based on FFTNet, a neural network for generating high quality audio waveform.

99, TITLE:      A Diffractive Neural Network with Weight-Noise-Injection Training
http://arxiv.org/abs/2006.04462
AUTHORS:        Jiashuo Shi
HIGHLIGHT:      We propose a diffractive neural network with strong robustness based on Weight Noise Injection training, which achieves accurate and fast optical-based classification while diffraction layers have a certain amount of surface shape error.

100, TITLE:     Continual Representation Learning for Biometric Identification
http://arxiv.org/abs/2006.04455
AUTHORS:        Bo Zhao ; Shixiang Tang ; Dapeng Chen ; Hakan Bilen ; Rui Zhao
HIGHLIGHT:      In this paper, we propose a new continual learning (CL) setting, namely ``continual representation learning'', which focuses on learning better representation in a continuous way. We also provide two large-scale multi-step benchmarks for biometric identification, where the visual appearance of different classes are highly relevant.

101, TITLE:     Novel Adaptive Binary Search Strategy-First Hybrid Pyramid- and Clustering-Based CNN Filter Pruning Method without Parameters Setting
http://arxiv.org/abs/2006.04451
AUTHORS:        Kuo-Liang Chung ; Yu-Lun Chang ; Bo-Wei Tsai
HIGHLIGHT:      In this paper, we propose an adaptive binary search-first hybrid pyramid- and clustering-based (ABSHPC-based) method for pruning filters automatically.

102, TITLE:     On Universalized Adversarial and Invariant Perturbations
http://arxiv.org/abs/2006.04449
AUTHORS:        Sandesh Kamath ; Amit Deshpande ; K V Subrahmanyam
COMMENTS:       Some part of this work was presented in ICML 2018 Workshop on "Towards learning with limited labels: Equivariance, Invariance,and Beyond" as "Understanding Adversarial Robustness of Symmetric Networks"
HIGHLIGHT:      In this paper, we study the effectiveness of SVD-Universal on GCNNs as they gain rotation invariance through higher degree of training augmentation.

103, TITLE:     Liquid Time-constant Networks
http://arxiv.org/abs/2006.04439
AUTHORS:        Ramin Hasani ; Mathias Lechner ; Alexander Amini ; Daniela Rus ; Radu Grosu
COMMENTS:       32 pages, 9 figures
HIGHLIGHT:      We introduce a new class of time-continuous recurrent neural network models.

104, TITLE:     CAST: A Correlation-based Adaptive Spectral Clustering Algorithm on Multi-scale Data
http://arxiv.org/abs/2006.04435
AUTHORS:        Xiang Li ; Ben Kao ; Caihua Shan ; Dawei Yin ; Martin Ester
HIGHLIGHT:      We propose the algorithm CAST that applies trace Lasso to regularize the coefficient matrix.

105, TITLE:     Training Deep Spiking Neural Networks
http://arxiv.org/abs/2006.04436
AUTHORS:        Eimantas Ledinauskas ; Julius Ruseckas ; Alfonsas Juršėnas ; Giedrius Buračas
HIGHLIGHT:      In this work we directly train deep SNNs using backpropagation with surrogate gradient and find that due to implicitly recurrent nature of feed forward SNN's the exploding or vanishing gradient problem severely hinders their training.

106, TITLE:     Visual Prediction of Priors for Articulated Object Interaction
http://arxiv.org/abs/2006.03979
AUTHORS:        Caris Moses ; Michael Noseworthy ; Leslie Pack Kaelbling ; Tomás Lozano-Pérez ; Nicholas Roy
COMMENTS:       IEEE International Conference on Robotics and Automation (ICRA) 2020
HIGHLIGHT:      We develop a method, Contextual Prior Prediction, which provides a means of transferring knowledge between interactions in similar domains through vision.

107, TITLE:     Generative Adversarial Phonology: Modeling unsupervised phonetic and phonological learning with neural networks
http://arxiv.org/abs/2006.03965
AUTHORS:        Gašper Beguš
COMMENTS:       Provisionally accepted in Frontiers in Artificial Intelligence
HIGHLIGHT:      The paper proposes a technique for establishing the network's internal representations that identifies latent variables that correspond to, for example, presence of [s] and its spectral properties.

108, TITLE:     Detecting Emergent Intersectional Biases: Contextualized Word Embeddings Contain a Distribution of Human-like Biases
http://arxiv.org/abs/2006.03955
AUTHORS:        Wei Guo ; Aylin Caliskan
COMMENTS:       11 pages, 1 figure, 1 table
HIGHLIGHT:      We propose a new comprehensive method, Contextualized Embedding Association Test (CEAT), based on the distribution of 10,000 pooled effect magnitudes of bias in embedding variations and a random-effects model, dispensing with templates.

109, TITLE:     ValNorm: A New Word Embedding Intrinsic Evaluation Method Reveals Valence Biases are Consistent Across Languages and Over Decades
http://arxiv.org/abs/2006.03950
AUTHORS:        Autumn Toney ; Aylin Caliskan
COMMENTS:       11 pages, 3 figures, 5 tables
HIGHLIGHT:      By extending methods that quantify human-like biases in word embeddings, we introduce ValNorm, a new word embedding intrinsic evaluation task, and the first unsupervised method that estimates the affective meaning of valence in words with high accuracy.

110, TITLE:     Self-Supervised Dynamic Networks for Covariate Shift Robustness
http://arxiv.org/abs/2006.03952
AUTHORS:        Tomer Cohen ; Noy Shulman ; Hai Morgenstern ; Roey Mechrez ; Erez Farhan
HIGHLIGHT:      In this work, we propose Self-Supervised Dynamic Networks (SSDN): an input-dependent mechanism, inspired by dynamic networks, that allows a self-supervised network to predict the weights of the main network, and thus directly handle covariate shifts at test-time.

111, TITLE:     The Convergence Indicator: Improved and completely characterized parameter bounds for actual convergence of Particle Swarm Optimization
http://arxiv.org/abs/2006.03944
AUTHORS:        Bernd Bassimir ; Alexander Raß ; Rolf Wanka
HIGHLIGHT:      In this paper we focus on the convergence of the particle swarm, i.e., the exploitation phase of the algorithm.

112, TITLE:     An Efficient $k$-modes Algorithm for Clustering Categorical Datasets
http://arxiv.org/abs/2006.03936
AUTHORS:        Karin S. Dorman ; Ranjan Maitra
COMMENTS:       28 pages, 16 figures, 5 tables
HIGHLIGHT:      We provide a fast and computationally efficient implementation of $k$-modes, OTQT, and prove that it can find superior clusterings to existing algorithms.

113, TITLE:     Robust watermarking with double detector-discriminator approach
http://arxiv.org/abs/2006.03921
AUTHORS:        Marcin Plata ; Piotr Syga
HIGHLIGHT:      In this paper we present a novel deep framework for a watermarking - a technique of embedding a transparent message into an image in a way that allows retrieving the message from a (perturbed) copy, so that copyright infringement can be tracked.

114, TITLE:     Self-supervising Fine-grained Region Similarities for Large-scale Image Localization
http://arxiv.org/abs/2006.03926
AUTHORS:        Yixiao Ge ; Haibo Wang ; Feng Zhu ; Rui Zhao ; Hongsheng Li
HIGHLIGHT:      To tackle this challenge, we propose to self-supervise image-to-region similarities in order to fully explore the potential of difficult positive images alongside their sub-regions.

115, TITLE:     A Robust Attentional Framework for License Plate Recognition in the Wild
http://arxiv.org/abs/2006.03919
AUTHORS:        Linjiang Zhang ; Peng Wang ; Hui Li ; Zhen Li ; Chunhua Shen ; Yanning Zhang
HIGHLIGHT:      In this work, we propose a robust framework for license plate recognition in the wild. Moreover, we released a new license plate dataset, named "CLPD", with 1200 images from all 31 provinces in mainland China.

116, TITLE:     Bitcoin covenants unchained
http://arxiv.org/abs/2006.03918
AUTHORS:        Massimo Bartoletti ; Stefano Lande ; Roberto Zunino
HIGHLIGHT:      In this paper we propose a formal model of covenants, which can be implemented with minor modifications to Bitcoin.

117, TITLE:     Complexity for deep neural networks and other characteristics of deep feature representations
http://arxiv.org/abs/2006.04791
AUTHORS:        Romuald A. Janik ; Przemek Witaszczyk
COMMENTS:       8 pages + 6 supplementary pages, 7+7 figures
HIGHLIGHT:      We define a notion of complexity, motivated by considerations of circuit complexity, which quantifies the nonlinearity of the computation of a neural network, as well as a complementary measure of the effective dimension of feature representations.

118, TITLE:     Are We Hungry for 3D LiDAR Data for Semantic Segmentation?
http://arxiv.org/abs/2006.04307
AUTHORS:        Biao Gao ; Yancheng Pan ; Chengkun Li ; Sibo Geng ; Huijing Zhao
COMMENTS:       26 pages, 18 figures
HIGHLIGHT:      This survey aims to explore whether and how we are hungry for 3D LiDAR data for semantic segmentation.

119, TITLE:     Text Detection and Recognition in the Wild: A Review
http://arxiv.org/abs/2006.04305
AUTHORS:        Zobeir Raisi ; Mohamed A. Naiel ; Paul Fieguth ; Steven Wardell ; John Zelek
HIGHLIGHT:      Thus, unlike previous surveys in this field, the objectives of this survey are as follows: first, offering the reader not only a review on the recent advancement in scene text detection and recognition, but also presenting the results of conducting extensive experiments using a unified evaluation framework that assesses pre-trained models of the selected methods on challenging cases, and applies the same evaluation criteria on these techniques.

120, TITLE:     What's the Difference Between Professional Human and Machine Translation? A Blind Multi-language Study on Domain-specific MT
http://arxiv.org/abs/2006.04781
AUTHORS:        Lukas Fischer ; Samuel Läubli
COMMENTS:       EAMT 2020 (Research Track)
HIGHLIGHT:      We find that the post-editing effort for MT segments is only higher in two out of three language pairs, and that the number of segments with wrong terminology, omissions, and typographical problems is similar in HT.

121, TITLE:     Fair Classification with Noisy Protected Attributes
http://arxiv.org/abs/2006.04778
AUTHORS:        L. Elisa Celis ; Lingxiao Huang ; Nisheeth K. Vishnoi
HIGHLIGHT:      We propose a ``denoised'' fair optimization formulation that can incorporate very general fairness goals via a set of constraints, mitigates the effects of such noise perturbations, and comes with provable guarantees.

122, TITLE:     An Astrocyte-Modulated Neuromorphic Central Pattern Generator for Hexapod Robot Locomotion on Intel's Loihi
http://arxiv.org/abs/2006.04765
AUTHORS:        Ioannis Polykretis ; Konstantinos P. Michmizos
COMMENTS:       8 pages, 7 figures, International Conference on Neuromorphic Systems (ICONS) 2020
HIGHLIGHT:      Here, we propose a brain-morphic CPG controler based on a comprehensive spiking neural-astrocytic network that generates two gait patterns for a hexapod robot.

123, TITLE:     Motion Prediction using Trajectory Sets and Self-Driving Domain Knowledge
http://arxiv.org/abs/2006.04767
AUTHORS:        Freddy A. Boulton ; Elena Corina Grigore ; Eric M. Wolff
HIGHLIGHT:      Our final contribution is a detailed comparison of classification and ordinal regression on two public self-driving datasets.

124, TITLE:     A Heuristically Self-Organised Linguistic Attribute Deep Learning in Edge Computing For IoT Intelligence
http://arxiv.org/abs/2006.04766
AUTHORS:        Hongmei He ; Zhenhuan Zhu
HIGHLIGHT:      In this paper, we propose a heuristic approach to constructing an LAH, embedded with LDTs for decision making or classification by utilising the distance correlations between attributes and between attributes and the goal variable.

125, TITLE:     Outlier Detection Using a Novel method: Quantum Clustering
http://arxiv.org/abs/2006.04760
AUTHORS:        Ding Liu ; Hui Li
COMMENTS:       9 pages, 18 figures
HIGHLIGHT:      We propose a new assumption in outlier detection: Normal data instances are commonly located in the area that there is hardly any fluctuation on data density, while outliers are often appeared in the area that there is violent fluctuation on data density.

126, TITLE:     Language Modeling for Formal Mathematics
http://arxiv.org/abs/2006.04757
AUTHORS:        Markus N. Rabe ; Dennis Lee ; Kshitij Bansal ; Christian Szegedy
HIGHLIGHT:      To train language models for formal mathematics, we propose a novel skip-tree task, which outperforms standard language modeling tasks on our reasoning benchmarks.

127, TITLE:     Approximate learning of high dimensional Bayesian network structures via pruning of Candidate Parent Sets
http://arxiv.org/abs/2006.04753
AUTHORS:        Zhigao Guo ; Anthony C. Constantinou
HIGHLIGHT:      Score-based algorithms that learn Bayesian Network (BN) structures provide solutions ranging from different levels of approximate learning to exact learning.

128, TITLE:     The Golden Ratio of Learning and Momentum
http://arxiv.org/abs/2006.04751
AUTHORS:        Stefan Jaeger
COMMENTS:       10 pages, 3 figures, under review
HIGHLIGHT:      This paper proposes a new information-theoretical loss function motivated by neural signal processing in a synapse.

129, TITLE:     The Criminality From Face Illusion
http://arxiv.org/abs/2006.03895
AUTHORS:        Kevin W. Bowyer ; Michael King ; Walter Scheirer
HIGHLIGHT:      Predicting criminality from face may initially seem similar to other facial analytics, but we argue that attempts to create a criminality-from-face algorithm are necessarily doomed to fail, that apparently promising experimental results in recent publications are an illusion resulting from inadequate experimental design, and that there is potentially a large social cost to belief in the criminality from face illusion.

130, TITLE:     Ensemble Network for Ranking Images Based on Visual Appeal
http://arxiv.org/abs/2006.03898
AUTHORS:        Sachin Singh ; Victor Sanchez ; Tanaya Guha
HIGHLIGHT:      We propose a computational framework for ranking images (group photos in particular) taken at the same event within a short time span. Owing to the unavailability of suitable databases, we created a new database of manually annotated group photos taken during various social events.

131, TITLE:     Unsupervised Transfer Learning with Self-Supervised Remedy
http://arxiv.org/abs/2006.04737
AUTHORS:        Jiabo Huang ; Shaogang Gong
HIGHLIGHT:      In this work, we address this problem by transfer clustering that aims to learn a discriminative latent space of the unlabelled target data in a novel domain by knowledge transfer from labelled related domains.

132, TITLE:     Reinforcement Learning Under Moral Uncertainty
http://arxiv.org/abs/2006.04734
AUTHORS:        Adrien Ecoffet ; Joel Lehman
COMMENTS:       33 pages, 17 figures
HIGHLIGHT:      Inspired by such work, this paper proposes a formalism that translates such insights to the field of reinforcement learning.

133, TITLE:     Biomechanics-informed Neural Networks for Myocardial Motion Tracking in MRI
http://arxiv.org/abs/2006.04725
AUTHORS:        Chen Qin ; Shuo Wang ; Chen Chen ; Huaqi Qiu ; Wenjia Bai ; Daniel Rueckert
COMMENTS:       The paper is early accepted to MICCAI 2020
HIGHLIGHT:      In contrast to most of the current approaches which impose explicit regularisation terms such as smoothness, in this paper we propose a novel method that can implicitly learn biomechanics-informed regularisation.

134, TITLE:     Scalene: Scripting-Language Aware Profiling for Python
http://arxiv.org/abs/2006.03879
AUTHORS:        Emery D. Berger
HIGHLIGHT:      This paper introduces scripting-language aware profiling, and presents Scalene, an implementation of scripting-language aware profiling for Python.

135, TITLE:     Modeling Discourse Structure for Document-level Neural Machine Translation
http://arxiv.org/abs/2006.04721
AUTHORS:        Junxuan Chen ; Xiang Li ; Jiarui Zhang ; Chulun Zhou ; Jianwei Cui ; Bin Wang ; Jinsong Su
HIGHLIGHT:      In this paper, we propose to improve document-level NMT with the aid of discourse structure information.

136, TITLE:     ARID: A New Dataset for Recognizing Action in the Dark
http://arxiv.org/abs/2006.03876
AUTHORS:        Yuecong Xu ; Jianfei Yang ; Haozhi Cao ; Kezhi Mao ; Jianxiong Yin ; Simon See
COMMENTS:       6 pages, 7 figures
HIGHLIGHT:      In this paper, we explored the task of action recognition in dark videos.

137, TITLE:     Towards large-scale, automated, accurate detection of CCTV camera objects using computer vision. Applications and implications for privacy, safety, and cybersecurity. (Preprint)
http://arxiv.org/abs/2006.03870
AUTHORS:        Hannu Turtiainen ; Andrei Costin ; Timo Hamalainen ; Tuomo Lahtinen
HIGHLIGHT:      To close this gap, with this paper we introduce the first and only computer vision MS COCO-compatible models that are able to accurately detect CCTV and video surveillance cameras in images and video frames.

138, TITLE:     A Cross-Task Analysis of Text Span Representations
http://arxiv.org/abs/2006.03866
AUTHORS:        Shubham Toshniwal ; Haoyue Shi ; Bowen Shi ; Lingyu Gao ; Karen Livescu ; Kevin Gimpel
COMMENTS:       RepL4NLP 2020
HIGHLIGHT:      In this paper, we conduct a comprehensive empirical evaluation of six span representation methods using eight pretrained language representation models across six tasks, including two tasks that we introduce.

139, TITLE:     Energy Constraints Improve Liquid State Machine Performance
http://arxiv.org/abs/2006.04716
AUTHORS:        Andrew Fountain ; Cory Merkel
COMMENTS:       8 pages, 5 figures. Submitted to ICONS 2020
HIGHLIGHT:      A model of metabolic energy constraints is applied to a liquid state machine in order to analyze its effects on network performance.

140, TITLE:     ResKD: Residual-Guided Knowledge Distillation
http://arxiv.org/abs/2006.04719
AUTHORS:        Xuewei Li ; Songyuan Li ; Bourahla Omar ; Xi Li
HIGHLIGHT:      In this paper, we see knowledge distillation in a fresh light, using the knowledge gap between a teacher and a student as guidance to train a lighter-weight student called res-student.

141, TITLE:     Cross-Domain Segmentation with Adversarial Loss and Covariate Shift for Biomedical Imaging
http://arxiv.org/abs/2006.04390
AUTHORS:        Bora Baydar ; Savas Ozkan ; A. Emre Kavur ; N. Sinem Gezer ; M. Alper Selver ; Gozde Bozdagi Akar
HIGHLIGHT:      For this purpose, this manuscript aims to implement a novel model that can learn robust representations from cross-domain data by encapsulating distinct and shared patterns from different modalities.

142, TITLE:     Learning the Compositional Visual Coherence for Complementary Recommendations
http://arxiv.org/abs/2006.04380
AUTHORS:        Zhi Li ; Bo Wu ; Qi Liu ; Likang Wu ; Hongke Zhao ; Tao Mei
COMMENTS:       Early version accepted by IJCAI2020
HIGHLIGHT:      Towards this end, in this paper, we propose a novel Content Attentive Neural Network (CANN) to model the comprehensive compositional coherence on both global contents and semantic contents.

143, TITLE:     An ASP approach for reasoning in a concept-aware multipreferential lightweight DL
http://arxiv.org/abs/2006.04387
AUTHORS:        Laura Giordano ; Daniele Theseider Dupré
COMMENTS:       23 pages
HIGHLIGHT:      In this paper we develop a concept aware multi-preferential semantics for dealing with typicality in description logics, where preferences are associated with concepts, starting from a collection of ranked TBoxes containing defeasible concept inclusions.

144, TITLE:     Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection
http://arxiv.org/abs/2006.04388
AUTHORS:        Xiang Li ; Wenhai Wang ; Lijun Wu ; Shuo Chen ; Xiaolin Hu ; Jun Li ; Jinhui Tang ; Jian Yang
COMMENTS:       14 pages, 14 figures
HIGHLIGHT:      To address the problems, we design new representations for these elements.

145, TITLE:     Speaker Diarization as a Fully Online Learning Problem in MiniVox
http://arxiv.org/abs/2006.04376
AUTHORS:        Baihan Lin ; Xinxin Zhang
HIGHLIGHT:      We proposed a novel AI framework to conduct real-time multi-speaker diarization and recognition without prior registration and pretraining in a fully online learning setting. First, we proposed a new benchmark to evaluate the rarely studied fully online speaker diarization problem.

146, TITLE:     Semantics-Driven Unsupervised Learning for Monocular Depth and Ego-Motion Estimation
http://arxiv.org/abs/2006.04371
AUTHORS:        Xiaobin Wei ; Jianjiang Feng ; Jie Zhou
HIGHLIGHT:      We propose a semantics-driven unsupervised learning approach for monocular depth and ego-motion estimation from videos in this paper.

147, TITLE:     Photoacoustic Microscopy with Sparse Data Enabled by Convolutional Neural Networks for Fast Imaging
http://arxiv.org/abs/2006.04368
AUTHORS:        Jiasheng Zhou ; Da He ; Xiaoyu Shang ; Zhendong Guo ; Sung-liang Chen ; Jiajia Luo
COMMENTS:       13 pages (including 2 pages of supplementary materials)
HIGHLIGHT:      In this work, we propose a method using convolutional neural networks (CNNs) to improve the quality of sparse PAM images, thereby speeding up image acquisition while keeping good image quality.

148, TITLE:     Hallucinating Value: A Pitfall of Dyna-style Planning with Imperfect Environment Models
http://arxiv.org/abs/2006.04363
AUTHORS:        Taher Jafferjee ; Ehsan Imani ; Erin Talvitie ; Martha White ; Micheal Bowling
COMMENTS:       9 pages, 7 figures,
HIGHLIGHT:      In this paper, we investigate one type of model error: hallucinated states.

149, TITLE:     Neural Sparse Representation for Image Restoration
http://arxiv.org/abs/2006.04357
AUTHORS:        Yuchen Fan ; Jiahui Yu ; Yiqun Mei ; Yulun Zhang ; Yun Fu ; Ding Liu ; Thomas S. Huang
HIGHLIGHT:      Inspired by the robustness and efficiency of sparse representation in sparse coding based image restoration models, we investigate the sparsity of neurons in deep networks.

150, TITLE:     Associate-3Ddet: Perceptual-to-Conceptual Association for 3D Point Cloud Object Detection
http://arxiv.org/abs/2006.04356
AUTHORS:        Liang Du ; Xiaoqing Ye ; Xiao Tan ; Jianfeng Feng ; Zhenbo Xu ; Errui Ding ; Shilei Wen
COMMENTS:       8 pages, 5 figures, CVPR 2020
HIGHLIGHT:      In this paper, we innovatively propose a domain adaptation like approach to enhance the robustness of the feature representation.

151, TITLE:     Deep Neural Network Based Real-time Kiwi Fruit Flower Detection in an Orchard Environment
http://arxiv.org/abs/2006.04343
AUTHORS:        JongYoon Lim ; Ho Seok Ahn ; Mahla Nejati ; Jamie Bell ; Henry Williams ; Bruce A. MacDonald
COMMENTS:       ACRA(Australian Robotics and Automation Association) 2019
HIGHLIGHT:      In this paper, we present a novel approach to kiwi fruit flower detection using Deep Neural Networks (DNNs) to build an accurate, fast, and robust autonomous pollination robot system.

152, TITLE:     Fast Synthetic LiDAR Rendering via Spherical UV Unwrapping of Equirectangular Z-Buffer Images
http://arxiv.org/abs/2006.04345
AUTHORS:        Mohammed Hossny ; Khaled Saleh ; Mohammed Attia ; Ahmed Abobakr ; Julie Iskander
COMMENTS:       8 pages and 6 figures. Copyright 2020. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by- nc-nd/4.0/
HIGHLIGHT:      In this paper, we present a novel method to simulate LiDAR point cloud with faster rendering time of 1 sec per frame.

153, TITLE:     PSPACE-completeness of Pulling Blocks to Reach a Goal
http://arxiv.org/abs/2006.04337
AUTHORS:        Joshua Ani ; Sualeh Asif ; Erik D. Demaine ; Yevhenii Diomidov ; Dylan Hendrickson ; Jayson Lynch ; Sarah Scheffler ; Adam Suhl
COMMENTS:       Full version of JCDCGGG2019 paper, 22 pages, 25 figures
HIGHLIGHT:      We prove PSPACE-completeness of all but one problem in a large space of pulling-block problems where the goal is for the agent to reach a target destination.

154, TITLE:     Randomized Policy Learning for Continuous State and Action MDPs
http://arxiv.org/abs/2006.04331
AUTHORS:        Hiteshi Sharma ; Rahul Jain
HIGHLIGHT:      We present \texttt{RANDPOL}, a generalized policy iteration algorithm for MDPs with continuous state and action spaces.

155, TITLE:     Characterizing Sociolinguistic Variation in the Competing Vaccination Communities
http://arxiv.org/abs/2006.04334
AUTHORS:        Shahan Ali Memon ; Aman Tyagi ; David R. Mortensen ; Kathleen M. Carley
COMMENTS:       11 pages, 4 tables, 1 figure, 1 algorithm, accepted to SBP-BRiMS 2020 -- International Conference on Social Computing, Behavioral-Cultural Modeling & Prediction and Behavior Representation in Modeling and Simulation
HIGHLIGHT:      Hence, in this paper, we conduct a sociolinguistic analysis of the two competing vaccination communities on Twitter: "pro-vaxxers" or individuals who believe in the effectiveness of vaccinations, and "anti-vaxxers" or individuals who are opposed to vaccinations.

156, TITLE:     Fully Convolutional Mesh Autoencoder using Efficient Spatially Varying Kernels
http://arxiv.org/abs/2006.04325
AUTHORS:        Yi Zhou ; Chenglei Wu ; Zimo Li ; Chen Cao ; Yuting Ye ; Jason Saragih ; Hao Li ; Yaser Sheikh
COMMENTS:       12 pages
HIGHLIGHT:      In this paper, we propose a non-template-specific fully convolutional mesh autoencoder for arbitrary registered mesh data.

157, TITLE:     Toward Scalable Algorithms for the Unsplittable Shortest Path Routing Problem
http://arxiv.org/abs/2006.04324
AUTHORS:        Amal Benhamiche ; Morgan Chopin
HIGHLIGHT:      In this paper, we consider the Delay Constrained Unsplittable Shortest Path Routing problem which arises in the field of traffic engineering for IP networks.

158, TITLE:     Ensemble Model with Batch Spectral Regularization and Data Blending for Cross-Domain Few-Shot Learning with Unlabeled Data
http://arxiv.org/abs/2006.04323
AUTHORS:        Zhen Zhao ; Bingyu Liu ; Yuhong Guo ; Jieping Ye
HIGHLIGHT:      We propose an Ensemble Model with Batch Spectral Regularization and Data Blending for the Track 2 of the CD-FSL challenge.

159, TITLE:     Counterfactual VQA: A Cause-Effect Look at Language Bias
http://arxiv.org/abs/2006.04315
AUTHORS:        Yulei Niu ; Kaihua Tang ; Hanwang Zhang ; Zhiwu Lu ; Xian-Sheng Hua ; Ji-Rong Wen
HIGHLIGHT:      In this paper, we propose a novel cause-effect look at the language bias, where the bias is formulated as the direct effect of question on answer from the view of causal inference.

160, TITLE:     Multi-step Estimation for Gradient-based Meta-learning
http://arxiv.org/abs/2006.04298
AUTHORS:        Jin-Hwa Kim ; Junyoung Park ; Yongseok Choi
COMMENTS:       17 pages, 5 figures
HIGHLIGHT:      To tackle this issue, we propose a simple yet straightforward method to reduce the cost by reusing the same gradient in a window of inner steps.

161, TITLE:     On the Maximum Cardinality Cut Problem in Proper Interval Graphs and Related Graph Classes
http://arxiv.org/abs/2006.03856
AUTHORS:        Arman Boyacı ; Tınaz Ekim ; Mordechai Shalom
HIGHLIGHT:      In this paper, we give FPT algorithms for the maximum cardinality cut problem in classes of graphs containing proper interval graphs and mixed unit interval graphs when parameterized by some new parameters that we introduce.

162, TITLE:     CycleGT: Unsupervised Graph-to-Text and Text-to-Graph Generation via Cycle Training
http://arxiv.org/abs/2006.04702
AUTHORS:        Qipeng Guo ; Zhijing Jin ; Xipeng Qiu ; Weinan Zhang ; David Wipf ; Zheng Zhang
COMMENTS:       Submitted to NeurIPS 2020
HIGHLIGHT:      We present CycleGT, an unsupervised training framework that can bootstrap from fully non-parallel graph and text datasets, iteratively back translate between the two forms, and use a novel pretraining strategy.

163, TITLE:     Instance segmentation of buildings using keypoints
http://arxiv.org/abs/2006.03858
AUTHORS:        Qingyu Li ; Lichao Mou ; Yuansheng Hua ; Yao Sun ; Pu Jin ; Yilei Shi ; Xiao Xiang Zhu
HIGHLIGHT:      In this paper, we propose a novel instance segmentation network for building segmentation in high-resolution remote sensing images.

164, TITLE:     EPARS: Early Prediction of At-risk Students with Online and Offline Learning Behaviors
http://arxiv.org/abs/2006.03857
AUTHORS:        Yu Yang ; Zhiyuan Wen ; Jiannong Cao ; Jiaxing Shen ; Hongzhi Yin ; Xiaofang Zhou
COMMENTS:       To be published in DASFAA 2020
HIGHLIGHT:      We propose a novel algorithm (EPARS) that could early predict STAR in a semester by modeling online and offline learning behaviors.

165, TITLE:     Multimodal Future Localization and Emergence Prediction for Objects in Egocentric View with a Reachability Prior
http://arxiv.org/abs/2006.04700
AUTHORS:        Osama Makansi ; Özgün Cicek ; Kevin Buchicchio ; Thomas Brox
COMMENTS:       In CVPR 2020
HIGHLIGHT:      In this paper, we investigate the problem of anticipating future dynamics, particularly the future location of other vehicles and pedestrians, in the view of a moving vehicle.

166, TITLE:     Principles to Practices for Responsible AI: Closing the Gap
http://arxiv.org/abs/2006.04707
AUTHORS:        Daniel Schiff ; Bogdana Rakova ; Aladdin Ayesh ; Anat Fanti ; Michael Lennon
COMMENTS:       Preprint draft. A version has been submitted to the 2020 European Conference on AI (ECAI) Workshop on "ADVANCING TOWARDS THE SDGs: ARTIFICIAL INTELLIGENCE FOR A FAIR, JUST AND EQUITABLE WORLD (AI4EQ)"
HIGHLIGHT:      This paper reviews the principles-to-practices gap.

167, TITLE:     Building a Heterogeneous, Large Scale Morphable Face Model
http://arxiv.org/abs/2006.03840
AUTHORS:        Claudio Ferrari ; Stefano Berretti ; Pietro Pala ; Alberto Del Bimbo
HIGHLIGHT:      In this manuscript, we present an approach that leverages a 3DMM to transfer its dense semantic annotation across a large set of heterogeneous 3D faces, establishing a dense correspondence between them.

168, TITLE:     Scaling Equilibrium Propagation to Deep ConvNets by Drastically Reducing its Gradient Estimator Bias
http://arxiv.org/abs/2006.03824
AUTHORS:        Axel Laborieux ; Maxence Ernoult ; Benjamin Scellier ; Yoshua Bengio ; Julie Grollier ; Damien Querlioz
HIGHLIGHT:      In this work, we show that a bias in the gradient estimate of EP, inherent in the use of finite nudging, is responsible for this phenomenon and that cancelling it allows training deep ConvNets by EP.

169, TITLE:     3D Self-Supervised Methods for Medical Imaging
http://arxiv.org/abs/2006.03829
AUTHORS:        Aiham Taleb ; Winfried Loetzsch ; Noel Danz ; Julius Severin ; Thomas Gaertner ; Benjamin Bergner ; Christoph Lippert
COMMENTS:       Submitted to NeurIPS 2020
HIGHLIGHT:      In this work, we leverage these techniques, and we propose 3D versions for five different self-supervised methods, in the form of proxy tasks.

170, TITLE:     An Empirical Analysis of the Impact of Data Augmentation on Knowledge Distillation
http://arxiv.org/abs/2006.03810
AUTHORS:        Deepan Das ; Haley Massa ; Abhimanyu Kulkarni ; Theodoros Rekatsinas
HIGHLIGHT:      In this work, we attempt to empirically analyse the impact of such augmentation strategies on the transfer of generalization between teacher and student models in a distillation setup.

171, TITLE:     UCLID-Net: Single View Reconstruction in Object Space
http://arxiv.org/abs/2006.03817
AUTHORS:        Benoit Guillard ; Edoardo Remelli ; Pascal Fua
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.

172, TITLE:     Extracting Cellular Location of Human Proteins Using Deep Learning
http://arxiv.org/abs/2006.03800
AUTHORS:        Hanke Chen
COMMENTS:       5 page, 10 figures
HIGHLIGHT:      To resolve this problem, we attempted to create an automatic image classifier using Machine Learning to locate human proteins with higher speed and accuracy than human beings.

173, TITLE:     Link Prediction for Temporally Consistent Networks
http://arxiv.org/abs/2006.03804
AUTHORS:        Mohamoud Ali ; Yugyung Lee ; Praveen Rao
HIGHLIGHT:      We propose a time-parameterized matrix (TP-matrix) and empirically demonstrate its effectiveness in non-bipartite, heterogeneous networks.

174, TITLE:     POLY-HOOT: Monte-Carlo Planning in Continuous Space MDPs with Non-Asymptotic Analysis
http://arxiv.org/abs/2006.04672
AUTHORS:        Weichao Mao ; Kaiqing Zhang ; Qiaomin Xie ; Tamer Başar
HIGHLIGHT:      In this paper, we consider Monte-Carlo planning in an environment with continuous state-action spaces, a much less understood problem with important applications in control and robotics.

175, TITLE:     Misinformation has High Perplexity
http://arxiv.org/abs/2006.04666
AUTHORS:        Nayeon Lee ; Yejin Bang ; Andrea Madotto ; Pascale Fung
HIGHLIGHT:      In this paper, we postulate that misinformation itself has higher perplexity compared to truthful statements, and propose to leverage the perplexity to debunk false claims in an unsupervised manner. We construct two new COVID-19-related test sets, one is scientific, and another is political in content, and empirically verify that our system performs favorably compared to existing systems. We are releasing these datasets publicly to encourage more research in debunking misinformation on COVID-19 and other topics.

176, TITLE:     MultiSpeech: Multi-Speaker Text to Speech with Transformer
http://arxiv.org/abs/2006.04664
AUTHORS:        Mingjian Chen ; Xu Tan ; Yi Ren ; Jin Xu ; Hao Sun ; Sheng Zhao ; Tao Qin
HIGHLIGHT:      In this paper, we develop a robust and high-quality multi-speaker Transformer TTS system called MultiSpeech, with several specially designed components/techniques to improve text-to-speech alignment: 1) a diagonal constraint on the weight matrix of encoder-decoder attention in both training and inference; 2) layer normalization on phoneme embedding in encoder to better preserve position information; 3) a bottleneck in decoder pre-net to prevent copy between consecutive speech frames.

177, TITLE:     Runtime Analysis of Evolutionary Algorithms via Symmetry Arguments
http://arxiv.org/abs/2006.04663
AUTHORS:        Benjamin Doerr
HIGHLIGHT:      We use an elementary argument building on group actions to prove that the selection-free steady state genetic algorithm analyzed by Sutton and Witt (GECCO 2019) takes an expected number of $\Omega(2^n / \sqrt n)$ iterations to find any particular target search point, regardless of the population size $\mu$.

178, TITLE:     Read what you need: Controllable Aspect-based Opinion Summarization of Tourist Reviews
http://arxiv.org/abs/2006.04660
AUTHORS:        Rajdeep Mukherjee ; Hari Chandana Peruri ; Uppada Vishnu ; Pawan Goyal ; Sourangshu Bhattacharya ; Niloy Ganguly
COMMENTS:       4 pages, accepted in the Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), 2020
HIGHLIGHT:      In this work, we argue the need and propose a solution for generating personalized aspect-based opinion summaries from large collections of online tourist reviews.

179, TITLE:     Semantic Graph-enhanced Visual Network for Zero-shot Learning
http://arxiv.org/abs/2006.04648
AUTHORS:        Yang Hu ; Guihua Wen ; Adriane Chapman ; Pei Yang ; Mingnan Luo ; Yingxue Xu ; Dan Dai ; Wendy Hall
COMMENTS:       15 pages, 8 figures
HIGHLIGHT:      In response to this, we propose the Graph-based Visual-Semantic Entanglement Network to conduct graph modeling of visual features, which is mapped to semantic attributes by using a knowledge graph, it contains several novel designs: 1.

180, TITLE:     Neural Architecture Search without Training
http://arxiv.org/abs/2006.04647
AUTHORS:        Joseph Mellor ; Jack Turner ; Amos Storkey ; Elliot J. Crowley
HIGHLIGHT:      In this work, we examine how the linear maps induced by data points correlate for untrained network architectures in the NAS-Bench-201 search space, and motivate how this can be used to give a measure of modelling flexibility which is highly indicative of a network's trained performance.

181, TITLE:     ColdGANs: Taming Language GANs with Cautious Sampling Strategies
http://arxiv.org/abs/2006.04643
AUTHORS:        Thomas Scialom ; Paul-Alexis Dray ; Sylvain Lamprier ; Benjamin Piwowarski ; Jacopo Staiano
HIGHLIGHT:      We propose to consider alternative exploration strategies in a GAN framework that we name ColdGANs, where we force the sampling to be close to the distribution modes to get smoother learning dynamics.

182, TITLE:     Deep Mining External Imperfect Data for Chest X-ray Disease Screening
http://arxiv.org/abs/2006.03796
AUTHORS:        Luyang Luo ; Lequan Yu ; Hao Chen ; Quande Liu ; Xi Wang ; Jiaqi Xu ; Pheng-Ann Heng
COMMENTS:       Accepted to IEEE Transactions on Medical Imaging
HIGHLIGHT:      In this paper, we argue that incorporating an external CXR dataset leads to imperfect training data, which raises the challenges.

183, TITLE:     Multi-Task Temporal Shift Attention Networks for On-Device Contactless Vitals Measurement
http://arxiv.org/abs/2006.03790
AUTHORS:        Xin Liu ; Josh Fromm ; Shwetak Patel ; Daniel McDuff
COMMENTS:       preprint
HIGHLIGHT:      We present a video-based and on-device optical cardiopulmonary vital sign measurement approach.

184, TITLE:     Learning to Play No-Press Diplomacy with Best Response Policy Iteration
http://arxiv.org/abs/2006.04635
AUTHORS:        Thomas Anthony ; Tom Eccles ; Andrea Tacchetti ; János Kramár ; Ian Gemp ; Thomas C. Hudson ; Nicolas Porcel ; Marc Lanctot ; Julien Pérolat ; Richard Everett ; Satinder Singh ; Thore Graepel ; Yoram Bachrach
HIGHLIGHT:      We propose a simple yet effective approximate best response operator, designed to handle large combinatorial action spaces and simultaneous moves.

185, TITLE:     No-Reference Image Quality Assessment via Feature Fusion and Multi-Task Learning
http://arxiv.org/abs/2006.03783
AUTHORS:        S. Alireza Golestaneh ; Kris Kitani
HIGHLIGHT:      In this paper, we propose a simple and yet effective general-purpose no-reference (NR) image quality assessment (IQA) framework based on multi-task learning.

186, TITLE:     MAGNet: Multi-Region Attention-Assisted Grounding of Natural Language Queries at Phrase Level
http://arxiv.org/abs/2006.03776
AUTHORS:        Amar Shrestha ; Krittaphat Pugdeethosapol ; Haowen Fang ; Qinru Qiu
COMMENTS:       Submitted to The 2020 European Conference on Computer Vision (ECCV 2020)
HIGHLIGHT:      Thus, we propose to utilize spatial attention networks for image-level visual-textual fusion preserving local (word) and global (phrase) information to refine region proposals with an in-network Region Proposal Network (RPN) and detect single or multiple regions for a phrase query.

187, TITLE:     Challenges and Thrills of Legal Arguments
http://arxiv.org/abs/2006.03773
AUTHORS:        Anurag Pallaprolu ; Radha Vaidya ; Aditya Swaroop Attawar
HIGHLIGHT:      We propose an extension, HumBERT, that attempts to perform continuous contextual argument generation using locally trained transformers.

188, TITLE:     A Comprehensive Survey on Aspect Based Sentiment Analysis
http://arxiv.org/abs/2006.04611
AUTHORS:        Kaustubh Yadav
HIGHLIGHT:      In this study, our aim is to explore the various methodologies practiced while performing ABSA, and providing a comparative study.

189, TITLE:     Deep Octree-based CNNs with Output-Guided Skip Connections for 3D Shape and Scene Completion
http://arxiv.org/abs/2006.03762
AUTHORS:        Peng-Shuai Wang ; Yang Liu ; Xin Tong
HIGHLIGHT:      We present a simple yet effective deep learning approach for completing the input noisy and incomplete shapes or scenes.

190, TITLE:     GRNet: Gridding Residual Network for Dense Point Cloud Completion
http://arxiv.org/abs/2006.03761
AUTHORS:        Haozhe Xie ; Hongxun Yao ; Shangchen Zhou ; Jiageng Mao ; Shengping Zhang ; Wenxiu Sun
HIGHLIGHT:      To solve this problem, we introduce 3D grids as intermediate representations to regularize unordered point clouds.

191, TITLE:     SoftFlow: Probabilistic Framework for Normalizing Flow on Manifolds
http://arxiv.org/abs/2006.04604
AUTHORS:        Hyeongju Kim ; Hyeonseung Lee ; Woo Hyun Kang ; Joun Yeop Lee ; Nam Soo Kim
COMMENTS:       17 pages, 15figures
HIGHLIGHT:      In this paper, we propose SoftFlow, a probabilistic framework for training normalizing flows on manifolds.

192, TITLE:     End-to-end learning for semiquantitative rating of COVID-19 severity on Chest X-rays
http://arxiv.org/abs/2006.04603
AUTHORS:        Alberto Signoroni ; Mattia Savardi ; Sergio Benini ; Nicola Adami ; Riccardo Leonardi ; Paolo Gibellini ; Filippo Vaccher ; Marco Ravanelli ; Andrea Borghesi ; Roberto Maroldi ; Davide Farina
COMMENTS:       22 pages, 9 figures, journal submission
HIGHLIGHT:      In this work we designed an end-to-end deep learning architecture for predicting, on Chest X-rays images (CRX), a multi-regional score conveying the degree of lung compromise in COVID-19 patients.

193, TITLE:     Expressivity of expand-and-sparsify representations
http://arxiv.org/abs/2006.03741
AUTHORS:        Sanjoy Dasgupta ; Christopher Tosh
HIGHLIGHT:      We study the benefits of this representation for subsequent learning.

194, TITLE:     Auxiliary Signal-Guided Knowledge Encoder-Decoder for Medical Report Generation
http://arxiv.org/abs/2006.03744
AUTHORS:        Mingjie Li ; Fuyu Wang ; Xiaojun Chang ; Xiaodan Liang
COMMENTS:       11 pages, 3 figures
HIGHLIGHT:      To address these problems, we propose an Auxiliary Signal-Guided Knowledge Encoder-Decoder (ASGK) to mimic radiologists' working patterns.

195, TITLE:     Simple Primary Colour Editing for Consumer Product Images
http://arxiv.org/abs/2006.03743
AUTHORS:        Han Gong ; Luwen Yu ; Stephen Westland
COMMENTS:       This is a working paper (pre-print). Code available at https://github.com/hangong/prod_recolor/
HIGHLIGHT:      We present a simple primary colour editing method for consumer product images.

196, TITLE:     Multi-Task Reinforcement Learning based Mobile Manipulation Control for Dynamic Object Tracking and Grasping
http://arxiv.org/abs/2006.04271
AUTHORS:        Cong Wang ; Qifeng Zhang ; Qiyan Tian ; Shuo Li ; Xiaohui Wang ; David Lane ; Yvan Petillot ; Ziyang Hong ; Sen Wang
COMMENTS:       6 pages, 7 figures, submitted to IROS2020
HIGHLIGHT:      In this paper, a multi-task reinforcement learning-based mobile manipulation control framework is proposed to achieve general dynamic object tracking and grasping.

197, TITLE:     EDropout: Energy-Based Dropout and Pruning of Deep Neural Networks
http://arxiv.org/abs/2006.04270
AUTHORS:        Hojjat Salehinejad ; Shahrokh Valaee
COMMENTS:       Submitted for peer-review
HIGHLIGHT:      We evaluate the proposed method on different flavours of ResNets, AlexNet, and SqueezeNet on the Kuzushiji, Fashion, CIFAR-10, CIFAR-100, and Flowers datasets, and compare with the state-of-the-art pruning and compression methods.

198, TITLE:     How useful is Active Learning for Image-based Plant Phenotyping?
http://arxiv.org/abs/2006.04255
AUTHORS:        Koushik Nagasubramanian ; Talukder Z. Jubery ; Fateme Fotouhi Ardakani ; Seyed Vahid Mirnezami ; Asheesh K. Singh ; Arti Singh ; Soumik Sarkar ; Baskar Ganapathysubramanian
HIGHLIGHT:      To overcome this challenge, active learning algorithms have been proposed that reduce the amount of labeling needed by deep learning models to achieve good predictive performance.

199, TITLE:     AdaLAM: Revisiting Handcrafted Outlier Detection
http://arxiv.org/abs/2006.04250
AUTHORS:        Luca Cavalli ; Viktor Larsson ; Martin Ralf Oswald ; Torsten Sattler ; Marc Pollefeys
HIGHLIGHT:      In this paper we revisit handcrafted approaches to outlier filtering.

200, TITLE:     Self-Representation Based Unsupervised Exemplar Selection in a Union of Subspaces
http://arxiv.org/abs/2006.04246
AUTHORS:        Chong You ; Chi Li ; Daniel P. Robinson ; Rene Vidal
COMMENTS:       In submission; conference version at ECCV'2018
HIGHLIGHT:      This paper proposes a new exemplar selection model that searches for a subset that best reconstructs all data points as measured by the $\ell_1$ norm of the representation coefficients.

201, TITLE:     Tensors over Semirings for Latent-Variable Weighted Logic Programs
http://arxiv.org/abs/2006.04232
AUTHORS:        Esma Balkir ; Daniel Gildea ; Shay Cohen
COMMENTS:       Accepted to IWPT
HIGHLIGHT:      In this paper we present a generalization of this concept: latent-variable semiring parsing.

202, TITLE:     Pre-training Polish Transformer-based Language Models at Scale
http://arxiv.org/abs/2006.04229
AUTHORS:        Sławomir Dadas ; Michał Perełkiewicz ; Rafał Poświata
HIGHLIGHT:      In this study, we present two language models for Polish based on the popular BERT architecture.

203, TITLE:     Unsupervised Learning for Subterranean Junction Recognition Based on 2D Point Cloud
http://arxiv.org/abs/2006.04225
AUTHORS:        Sina Sharif Mansouri ; Farhad Pourkamali-Anaraki ; Miguel Castano Arranz ; Ali-akbar Agha-mohammadi ; Joel Burdick ; George Nikolakopoulos
HIGHLIGHT:      This article proposes a novel unsupervised learning framework for detecting the number of tunnel junctions in subterranean environments based on acquired 2D point clouds.

204, TITLE:     AI-QMIX: Attention and Imagination for Dynamic Multi-Agent Reinforcement Learning
http://arxiv.org/abs/2006.04222
AUTHORS:        Shariq Iqbal ; Christian A. Schroeder de Witt ; Bei Peng ; Wendelin Böhmer ; Shimon Whiteson ; Fei Sha
HIGHLIGHT:      In this work, we tackle the problem of multi-agent reinforcement learning (MARL) in such dynamic scenarios.

205, TITLE:     Efficient Poverty Mapping using Deep Reinforcement Learning
http://arxiv.org/abs/2006.04224
AUTHORS:        Kumar Ayush ; Burak Uzkent ; Marshall Burke ; David Lobell ; Stefano Ermon
HIGHLIGHT:      To reduce acquisition costs while maintaining accuracy, we propose a reinforcement learning approach in which free low-resolution imagery is used to dynamically identify where to acquire costly high-resolution images, prior to performing a deep learning task on the high-resolution images.

206, TITLE:     Supervised Whole DAG Causal Discovery
http://arxiv.org/abs/2006.04697
AUTHORS:        Hebi Li ; Qi Xiao ; Jin Tian
HIGHLIGHT:      We propose a novel approach of modeling the whole DAG structure discovery as a supervised learning.

207, TITLE:     Unstructured Road Vanishing Point Detection Using the Convolutional Neural Network and Heatmap Regression
http://arxiv.org/abs/2006.04691
AUTHORS:        Yin-Bo Liu ; Ming Zeng ; Qing-Hao Meng
COMMENTS:       8 pages, 6 figures, under review
HIGHLIGHT:      In this paper, we proposed a novel solution combining the convolutional neural network (CNN) and heatmap regression to detect unstructured road VP.

208, TITLE:     Thoracic Disease Identification and Localization using Distance Learning and Region Verification
http://arxiv.org/abs/2006.04203
AUTHORS:        Cheng Zhang ; Francine Chen ; Yan-Ying Chen
HIGHLIGHT:      In this paper, we propose an alternative approach that learns discriminative features among triplets of images and cyclically trains on region features to verify whether attentive regions contain information indicative of a disease.

209, TITLE:     Graph Minors Meet Machine Learning: the Power of Obstructions
http://arxiv.org/abs/2006.04689
AUTHORS:        Faisal N. Abu-Khzam ; Mohamed Mahmoud Abd El-Wahab ; Noureldin Yosri
HIGHLIGHT:      In this paper we consider the effectiveness of selecting a proper training strategy by considering special problem instances called "obstructions" that we believe carry some intrinsic properties of the problem itself.

210, TITLE:     A Review of Emergency Incident Prediction, Resource Allocation and Dispatch Models
http://arxiv.org/abs/2006.04200
AUTHORS:        Ayan Mukhopadhyay ; Geoffrey Pettet ; Sayyed Vazirizade ; Yevgeniy Vorobeychik ; Mykel Kochenderfer ; Abhishek Dubey
HIGHLIGHT:      In this survey, we present models for incident prediction, resource allocation and dispatch concerning urban emergency incidents like accidents and crimes.

211, TITLE:     RBF Solver for Quaternions Interpolation
http://arxiv.org/abs/2006.04686
AUTHORS:        Rinaldi Fabio ; Dolci Daniele
COMMENTS:       23 pages, 21 figures, 2 txt files
HIGHLIGHT:      In this paper we adapt the RBF Solver to work with quaternions by taking advantage of their Lie Algebra and exponential map.

212, TITLE:     Learning Behaviors with Uncertain Human Feedback
http://arxiv.org/abs/2006.04201
AUTHORS:        Xu He ; Haipeng Chen ; Bo An
HIGHLIGHT:      To address this challenge, we introduce a novel feedback model that considers the uncertainty of human feedback.



==========Updates to Previous Papers==========
1, TITLE:       Earballs: Neural Transmodal Translation
http://arxiv.org/abs/2005.13291
AUTHORS:        Andrew Port ; Chelhwon Kim ; Mitesh Patel
COMMENTS:       9 pages, 3 figures
HIGHLIGHT:      This work describes a novel technique for leveraging machine learned feature embeddings to translate visual (and other types of) information into a perceptual audio domain, allowing users to perceive this information using only their aural faculty.

2, TITLE:       Using a min-cut generalisation to go beyond Boolean surjective VCSPs
http://arxiv.org/abs/1901.07107
AUTHORS:        Gregor Matl ; Stanislav Zivny
COMMENTS:       Full version of a STACS'19 paper
HIGHLIGHT:      We present a polynomial-time algorithm for enumerating all near-optimal solutions of this Bounded Generalised Min-Cut problem.

3, TITLE:       Elastic Boundary Projection for 3D Medical Image Segmentation
http://arxiv.org/abs/1812.00518
AUTHORS:        Tianwei Ni ; Lingxi Xie ; Huangjie Zheng ; Elliot K. Fishman ; Alan L. Yuille
COMMENTS:       Accepted to CVPR 2019
HIGHLIGHT:      In this paper, we bridge the gap between 2D and 3D using a novel approach named Elastic Boundary Projection (EBP).

4, TITLE:       Optimizing Loss Functions Through Multivariate Taylor Polynomial Parameterization
http://arxiv.org/abs/2002.00059
AUTHORS:        Santiago Gonzalez ; Risto Miikkulainen
HIGHLIGHT:      Whereas early work focused on genetic programming (GP) on tree representations, this paper proposes continuous CMA-ES optimization of multivariate Taylor polynomial parameterizations.

5, TITLE:       Multi-organ Segmentation over Partially Labeled Datasets with Multi-scale Feature Abstraction
http://arxiv.org/abs/2001.00208
AUTHORS:        Xi Fang ; Pingkun Yan
COMMENTS:       Accepted for publication at IEEE Transactions on Medical Imaging
HIGHLIGHT:      In this paper, we propose a unified training strategy that enables a novel multi-scale deep neural network to be trained on multiple partially labeled datasets for multi-organ segmentation.

6, TITLE:       Unsupervised Learning of Depth, Optical Flow and Pose with Occlusion from 3D Geometry
http://arxiv.org/abs/2003.00766
AUTHORS:        Guangming Wang ; Chi Zhang ; Hesheng Wang ; Jingchuan Wang ; Yong Wang ; Xinlei Wang
COMMENTS:       14 pages, 13 figures, under review
HIGHLIGHT:      Our experiments in KITTI dataset demonstrate that the model based on three regions, full and explicit segmentation of the occlusion region, the rigid region, and the non-rigid region with corresponding unsupervised losses can improve performance on three tasks significantly.

7, TITLE:       Causal Feature Learning for Utility-Maximizing Agents
http://arxiv.org/abs/2005.08792
AUTHORS:        David Kinney ; David Watson
HIGHLIGHT:      We propose a new technique, pragmatic causal feature learning (PCFL), which extends the original CFL algorithm in useful and intuitive ways.

8, TITLE:       Translating Natural Language Instructions for Behavioral Robot Navigation with a Multi-Head Attention Mechanism
http://arxiv.org/abs/2006.00697
AUTHORS:        Patricio Cerda-Mardini ; Vladimir Araujo ; Alvaro Soto
COMMENTS:       Accepted at ACL 2020 WiNLP workshop
HIGHLIGHT:      We propose a multi-head attention mechanism as a blending layer in a neural network model that translates natural language to a high level behavioral language for indoor robot navigation.

9, TITLE:       Analogical Proportions
http://arxiv.org/abs/2006.02854
AUTHORS:        Christian Antić
HIGHLIGHT:      This paper contributes to the foundations of artificial general intelligence by introducing an abstract algebraic framework of analogical proportions of the form `$a$ is to $b$ what $c$ is to $d$' in the general setting of universal algebra.

10, TITLE:      BHN: A Brain-like Heterogeneous Network
http://arxiv.org/abs/2005.12826
AUTHORS:        Tao Liu
COMMENTS:       Improve the readability, and add an image experiment
HIGHLIGHT:      Inspired by this, we propose a brain-like heterogeneous network (BHN), which can cooperatively learn a lot of distributed representations and one global attention representation.

11, TITLE:      A Simple Baseline to Semi-Supervised Domain Adaptation for Machine Translation
http://arxiv.org/abs/2001.08140
AUTHORS:        Di Jin ; Zhijing Jin ; Joey Tianyi Zhou ; Peter Szolovits
COMMENTS:       Under review
HIGHLIGHT:      In this work, we propose a simple but effect approach to the semi-supervised domain adaptation scenario of NMT, where the aim is to improve the performance of a translation model on the target domain consisting of only non-parallel data with the help of supervised source domain data.

12, TITLE:      ERNIE-GEN: An Enhanced Multi-Flow Pre-training and Fine-tuning Framework for Natural Language Generation
http://arxiv.org/abs/2001.11314
AUTHORS:        Dongling Xiao ; Han Zhang ; Yukun Li ; Yu Sun ; Hao Tian ; Hua Wu ; Haifeng Wang
COMMENTS:       The source codes and pre-trained models have been released at https://github.com/PaddlePaddle/ERNIE. We have also updated the performances of ERNIE-GEN under a larger scaled pre-training corpora in appendix A
HIGHLIGHT:      To address this issue, we propose an enhanced multi-flow sequence to sequence pre-training and fine-tuning framework named ERNIE-GEN, which bridges the discrepancy between training and inference with an infilling generation mechanism and a noise-aware generation method.

13, 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.

14, TITLE:      Feature-map-level Online Adversarial Knowledge Distillation
http://arxiv.org/abs/2002.01775
AUTHORS:        Inseop Chung ; SeongUk Park ; Jangho Kim ; Nojun Kwak
HIGHLIGHT:      Thus in this paper, we propose an online knowledge distillation method that transfers not only the knowledge of the class probabilities but also that of the feature map using the adversarial training framework.

15, TITLE:      Action for Better Prediction
http://arxiv.org/abs/2003.06082
AUTHORS:        Bernadette Bucher ; Karl Schmeckpeper ; Nikolai Matni ; Kostas Daniilidis
COMMENTS:       Additional visualizations of our results are available on our website at https://sites.google.com/view/action-for-better-prediction . Bernadette Bucher and Karl Schmeckpeper contributed equally
HIGHLIGHT:      We present an adversarial curiosity method which minimizes a score given by a discriminator network. We also present a validation dataset of action-conditioned video of robotic manipulation tasks on which we test the prediction performance of our trained models.

16, TITLE:      Coronavirus: Comparing COVID-19, SARS and MERS in the eyes of AI
http://arxiv.org/abs/2005.11524
AUTHORS:        Anas Tahir ; Yazan Qiblawey ; Amith Khandakar ; Tawsifur Rahman ; Uzair Khurshid ; Farayi Musharavati ; M. T. Islam ; Serkan Kiranyaz ; Muhammad E. H. Chowdhury
COMMENTS:       10 Figures, 4 Tables
HIGHLIGHT:      In this work, authors used deep machine learning algorithms along with innovative image pre-processing techniques to distinguish COVID-19 images from SARS and MERS images.

17, TITLE:      Character-level Japanese Text Generation with Attention Mechanism for Chest Radiography Diagnosis
http://arxiv.org/abs/2004.13846
AUTHORS:        Kenya Sakka ; Kotaro Nakayama ; Nisei Kimura ; Taiki Inoue ; Yusuke Iwasawa ; Ryohei Yamaguchi ; Yosimasa Kawazoe ; Kazuhiko Ohe ; Yutaka Matsuo
COMMENTS:       8 pages, 3 figures, 2 tables
HIGHLIGHT:      For deal with these two challenges, we proposed an end-to-end model that generates Japanese findings at the character-level from chest radiographs.

18, TITLE:      Template-based Question Answering using Recursive Neural Networks
http://arxiv.org/abs/2004.13843
AUTHORS:        Ram G Athreya ; Srividya Bansal ; Axel-Cyrille Ngonga-Ngomo ; Ricardo Usbeck
HIGHLIGHT:      We propose a neural network-based approach to automatically learn and classify natural language questions into its corresponding template using recursive neural networks.

19, TITLE:      Few-Shot Open-Set Recognition using Meta-Learning
http://arxiv.org/abs/2005.13713
AUTHORS:        Bo Liu ; Hao Kang ; Haoxiang Li ; Gang Hua ; Nuno Vasconcelos
HIGHLIGHT:      Few-Shot Open-Set Recognition using Meta-Learning

20, TITLE:      A Survey on Explainable Artificial Intelligence (XAI): Towards Medical XAI
http://arxiv.org/abs/1907.07374
AUTHORS:        Erico Tjoa ; Cuntai Guan
HIGHLIGHT:      We provide a review on interpretabilities suggested by different research works and categorize them.

21, TITLE:      Grafted network for person re-identification
http://arxiv.org/abs/2006.01967
AUTHORS:        Jiabao Wang ; Yang Li ; Shanshan Jiao ; Zhuang Miao ; Rui Zhang
HIGHLIGHT:      In order to relieve this problem, we propose a novel grafted network (GraftedNet), which is designed by grafting a high-accuracy rootstock and a light-weighted scion.

22, TITLE:      Restore from Restored: Single Image Denoising with Pseudo Clean Image
http://arxiv.org/abs/2003.04721
AUTHORS:        Seunghwan Lee ; Dongkyu Lee ; Donghyeon Cho ; Jiwon Kim ; Tae Hyun Kim
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:2001.02899
HIGHLIGHT:      Thus, we propose a new and effective denoising approach that can greatly outperform the state-of-the-art supervised/self-supervised denoising methods by adapting (fine-tuning) their network parameters to the given specific input through self-supervision without changing the original network architectures.

23, TITLE:      CDL: Curriculum Dual Learning for Emotion-Controllable Response Generation
http://arxiv.org/abs/2005.00329
AUTHORS:        Lei Shen ; Yang Feng
COMMENTS:       To appear at ACL 2020 (long paper)
HIGHLIGHT:      To alleviate these problems, we propose a novel framework named Curriculum Dual Learning (CDL) which extends the emotion-controllable response generation to a dual task to generate emotional responses and emotional queries alternatively.

24, TITLE:      Analysis of Random Perturbations for Robust Convolutional Neural Networks
http://arxiv.org/abs/2002.03080
AUTHORS:        Adam Dziedzic ; Sanjay Krishnan
HIGHLIGHT:      Recent work has extensively shown that randomized perturbations of neural networks can improve robustness to adversarial attacks.

25, TITLE:      Deep Fusion Siamese Network for Automatic Kinship Verification
http://arxiv.org/abs/2006.00143
AUTHORS:        Jun Yu ; Mengyan Li ; Xinlong Hao ; Guochen Xie
COMMENTS:       8 pages, 8 figures
HIGHLIGHT:      In this work, the challenging problem is progressively addressed in two respects.

26, TITLE:      Discriminative Dictionary Design for Action Classification in Still Images and Videos
http://arxiv.org/abs/2005.10149
AUTHORS:        Abhinaba Roy ; Biplab Banerjee ; Amir Hussain ; Soujanya Poria
HIGHLIGHT:      In this paper, we address the problem of action recognition from still images and videos.

27, TITLE:      M^3VSNet: Unsupervised Multi-metric Multi-view Stereo Network
http://arxiv.org/abs/2005.00363
AUTHORS:        Baichuan Huang ; Hongwei Yi ; Can Huang ; Yijia He ; Jingbin Liu ; Xiao Liu
COMMENTS:       The original top-level version is arXiv:2004.09722v2 but I upload the similar version to arXiv:2005.00363 mistakenly, which is overlapped with arXiv:2004.09722v2. So the submission is to make the two addresses keeping the same version
HIGHLIGHT:      In this paper, we propose a novel unsupervised multi-metric MVS network, named M^3VSNet, for dense point cloud reconstruction without any supervision.

28, TITLE:      FrequentNet : A New Interpretable Deep Learning Baseline for Image Classification
http://arxiv.org/abs/2001.01034
AUTHORS:        Yifei Li ; Zheng Wang ; Xiaoyu Lu ; Kuangyan Song ; Yiming Sun
HIGHLIGHT:      This paper has proposed a new baseline deep learning model of more benefits for image classification.

29, TITLE:      Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction
http://arxiv.org/abs/1906.03631
AUTHORS:        Osama Makansi ; Eddy Ilg ; Özgün Cicek ; Thomas Brox
COMMENTS:       In CVPR 2019
HIGHLIGHT:      In this work, we present an approach that involves the prediction of several samples of the future with a winner-takes-all loss and iterative grouping of samples to multiple modes.

30, TITLE:      Algorithms for $\ell_p$-based semi-supervised learning on graphs
http://arxiv.org/abs/1901.05031
AUTHORS:        Mauricio Flores Rios ; Jeff Calder ; Gilad Lerman
HIGHLIGHT:      We present several efficient and scalable algorithms for both the variational and game-theoretic formulations, and present numerical results on synthetic data and real data that illustrate the effectiveness of the $p$-Laplacian formulation for semi-supervised learning with few labels.

31, TITLE:      Emergence of Separable Manifolds in Deep Language Representations
http://arxiv.org/abs/2006.01095
AUTHORS:        Jonathan Mamou ; Hang Le ; Miguel Del Rio ; Cory Stephenson ; Hanlin Tang ; Yoon Kim ; SueYeon Chung
COMMENTS:       8 pages. 8 figures. Accepted to ICML 2020
HIGHLIGHT:      In this work, we utilize mean-field theoretic manifold analysis, a recent technique from computational neuroscience, to analyze the high dimensional geometry of language representations from large-scale contextual embedding models.

32, TITLE:      Line Art Correlation Matching Network for Automatic Animation Colorization
http://arxiv.org/abs/2004.06718
AUTHORS:        Zhang Qian ; Wang Bo ; Wen Wei ; Li Hai ; Liu Jun Hui
COMMENTS:       7pages,6 figures
HIGHLIGHT:      To address these challenges, we design a kind of correlation matching model (called CM) to align the colored reference in a learnable way and integrate the model into an U-Net structure generator in a coarse-to-fine manner.

33, TITLE:      Learning Multi-Agent Coordination through Graph-driven Communication
http://arxiv.org/abs/2002.05233
AUTHORS:        Emanuele Pesce ; Giovanni Montana
HIGHLIGHT:      A connectivity-driven communication (CDC) algorithm is proposed to address three key aspects: what agents to involve in the communication, what information content to share, and how often to share it.

34, TITLE:      DeepCOVIDExplainer: Explainable COVID-19 Diagnosis Based on Chest X-ray Images
http://arxiv.org/abs/2004.04582
AUTHORS:        Md. Rezaul Karim ; Till Döhmen ; Dietrich Rebholz-Schuhmann ; Stefan Decker ; Michael Cochez ; Oya Beyan
HIGHLIGHT:      In this paper, we propose an explainable deep neural networks(DNN)-based method for automatic detection of COVID-19 symptoms from CXR images, which we call DeepCOVIDExplainer.

35, TITLE:      Contextual Prediction Difference Analysis for Explaining Individual Image Classifications
http://arxiv.org/abs/1910.09086
AUTHORS:        Jindong Gu ; Volker Tresp
HIGHLIGHT:      Much effort has been devoted to understanding the decisions of deep neural networks in recent years.

36, TITLE:      Self-supervised classification of dynamic obstacles using the temporal information provided by videos
http://arxiv.org/abs/1910.09094
AUTHORS:        Sid Ali Hamideche ; Florent Chiaroni ; Mohamed-Cherif Rahal
HIGHLIGHT:      In this work, we propose to classify the detected obstacles depending on their motion pattern.

37, TITLE:      Adaptive Geo-Topological Independence Criterion
http://arxiv.org/abs/1810.02923
AUTHORS:        Baihan Lin ; Nikolaus Kriegeskorte
HIGHLIGHT:      We propose a class of adaptive (multi-threshold) test statistics, which form the basis for permutation tests.

38, TITLE:      Transfer Learning for Information Extraction with Limited Data
http://arxiv.org/abs/2003.03064
AUTHORS:        Minh-Tien Nguyen ; Viet-Anh Phan ; Le Thai Linh ; Nguyen Hong Son ; Le Tien Dung ; Miku Hirano ; Hajime Hotta
COMMENTS:       14 pages, 5 figures, PACLING conference
HIGHLIGHT:      This paper presents a practical approach to fine-grained information extraction.

39, TITLE:      Lightweight Residual Densely Connected Convolutional Neural Network
http://arxiv.org/abs/2001.00526
AUTHORS:        Fahimeh Fooladgar ; Shohreh Kasaei
HIGHLIGHT:      In this paper, the lightweight residual densely connected blocks are proposed to guaranty the deep supervision, efficient gradient flow, and feature reuse abilities of convolutional neural network.

40, TITLE:      Reservoir computing approaches for representation and classification of multivariate time series
http://arxiv.org/abs/1803.07870
AUTHORS:        Filippo Maria Bianchi ; Simone Scardapane ; Sigurd Løkse ; Robert Jenssen
HIGHLIGHT:      In this paper we introduce the reservoir model space, an unsupervised approach based on RC to learn vectorial representations of MTS.

41, TITLE:      DAugNet: Unsupervised, Multi-source, Multi-target, and Life-long Domain Adaptation for Semantic Segmentation of Satellite Images
http://arxiv.org/abs/2005.06216
AUTHORS:        Onur Tasar ; Alain Giros ; Yuliya Tarabalka ; Pierre Alliez ; Sébastien Clerc
HIGHLIGHT:      We propose a novel approach, coined DAugNet, for unsupervised, multi-source, multi-target, and life-long domain adaptation of satellite images.

42, TITLE:      Complexity of controlled bad sequences over finite sets of $\mathbb{N}^d$
http://arxiv.org/abs/1909.01667
AUTHORS:        A. R. Balasubramanian
HIGHLIGHT:      We provide upper and lower bounds for the length of controlled bad sequences over the majoring and the minoring orderings of finite sets of $\mathbb{N}^d$.

43, TITLE:      Principal Neighbourhood Aggregation for Graph Nets
http://arxiv.org/abs/2004.05718
AUTHORS:        Gabriele Corso ; Luca Cavalleri ; Dominique Beaini ; Pietro Liò ; Petar Veličković
HIGHLIGHT:      Accordingly, we propose Principal Neighbourhood Aggregation (PNA), a novel architecture combining multiple aggregators with degree-scalers (which generalize the sum aggregator).

44, TITLE:      Adversarial Ranking Attack and Defense
http://arxiv.org/abs/2002.11293
AUTHORS:        Mo Zhou ; Zhenxing Niu ; Le Wang ; Qilin Zhang ; Gang Hua
HIGHLIGHT:      In this paper, we propose two attacks against deep ranking systems, i.e., Candidate Attack and Query Attack, that can raise or lower the rank of chosen candidates by adversarial perturbations.

45, TITLE:      Rule Extraction in Unsupervised Anomaly Detection for Model Explainability: Application to OneClass SVM
http://arxiv.org/abs/1911.09315
AUTHORS:        Alberto Barbado ; Óscar Corcho ; Richard Benjamins
COMMENTS:       23 pages, 18 figures
HIGHLIGHT:      Together with that, we propose algorithms to compute metrics related with XAI regarding the "comprehensivility", "representativeness", "stability" and "diversity" of the rules extracted.

46, TITLE:      Towards Understanding the Adversarial Vulnerability of Skeleton-based Action Recognition
http://arxiv.org/abs/2005.07151
AUTHORS:        Tianhang Zheng ; Sheng Liu ; Changyou Chen ; Junsong Yuan ; Baochun Li ; Kui Ren
HIGHLIGHT:      In this paper, we attempt to conduct a thorough study towards understanding the adversarial vulnerability of skeleton-based action recognition.

47, TITLE:      AeGAN: Time-Frequency Speech Denoising via Generative Adversarial Networks
http://arxiv.org/abs/1910.12620
AUTHORS:        Sherif Abdulatif ; Karim Armanious ; Karim Guirguis ; Jayasankar T. Sajeev ; Bin Yang
COMMENTS:       5 pages, 4 figures and 2 Tables. Accepted in EUSIPCO 2020
HIGHLIGHT:      In this paper, a generative adversarial network (GAN) based framework is investigated for the task of speech enhancement, more specifically speech denoising of audio tracks.

48, TITLE:      RetinaMask: A Face Mask detector
http://arxiv.org/abs/2005.03950
AUTHORS:        Mingjie Jiang ; Xinqi Fan ; Hong Yan
HIGHLIGHT:      In this paper, we propose RetinaFaceMask, which is a high-accuracy and efficient face mask detector.

49, TITLE:      Multistage s-t Path: Confronting Similarity with Dissimilarity
http://arxiv.org/abs/2002.07569
AUTHORS:        Till Fluschnik ; Rolf Niedermeier ; Carsten Schubert ; Philipp Zschoche
HIGHLIGHT:      Motivated by this fact and natural applications of this scenario e.g. in traffic route planning, we perform a parameterized complexity analysis.

50, TITLE:      Advances in Deep Space Exploration via Simulators & Deep Learning
http://arxiv.org/abs/2002.04051
AUTHORS:        James Bird ; Linda Petzold ; Philip Lubin ; Julia Deacon
COMMENTS:       16 pages, 7 figures, 2 tables, currently under review
HIGHLIGHT:      We introduce simulator-based methods that leverage artificial intelligence, mostly in the form of computer vision, in order to solve all three of these issues.

51, TITLE:      Cross-modal Learning for Multi-modal Video Categorization
http://arxiv.org/abs/2003.03501
AUTHORS:        Palash Goyal ; Saurabh Sahu ; Shalini Ghosh ; Chul Lee
HIGHLIGHT:      In this paper, we focus on the problem of video categorization using a multi-modal ML technique.

52, TITLE:      Milking CowMask for Semi-Supervised Image Classification
http://arxiv.org/abs/2003.12022
AUTHORS:        Geoff French ; Avital Oliver ; Tim Salimans
COMMENTS:       11 pages, 2 figures, submitted to NeurIPS 2020
HIGHLIGHT:      Here, we present a novel mask-based augmentation method called CowMask.

53, TITLE:      Source Camera Verification from Strongly Stabilized Videos
http://arxiv.org/abs/1912.05018
AUTHORS:        Enes Altinisik ; Husrev Taha Sencar
HIGHLIGHT:      To address this challenge, we introduce a source camera verification method for videos that takes into account the spatially variant nature of stabilization transformations and assumes a larger degree of freedom in their search.

54, TITLE:      Auditing and Achieving Intersectional Fairness in Classification Problems
http://arxiv.org/abs/1911.01468
AUTHORS:        Giulio Morina ; Viktoriia Oliinyk ; Julian Waton ; Ines Marusic ; Konstantinos Georgatzis
HIGHLIGHT:      We present a comprehensive framework for auditing and achieving intersectional fairness in classification problems: we define a suite of metrics to assess intersectional fairness in the data or model outputs by extending known single-attribute fairness metrics, and propose methods for robustly estimating them even when some intersectional subgroups are underrepresented.

55, TITLE:      Isotropic Maximization Loss and Entropic Score: Accurate, Fast, Scalable, Turnkey, and Native Neural Networks Out-of-Distribution Detection
http://arxiv.org/abs/1908.05569
AUTHORS:        David Macêdo ; Tsang Ing Ren ; Cleber Zanchettin ; Adriano L. I. Oliveira ; Teresa Ludermir
HIGHLIGHT:      In this paper, we argue that the low ODD performance of neural networks is mainly due to SoftMax loss anisotropy.

56, TITLE:      Out-of-Vocabulary Embedding Imputation with Grounded Language Information by Graph Convolutional Networks
http://arxiv.org/abs/1906.03753
AUTHORS:        Ziyi Yang ; Chenguang Zhu ; Vin Sachidananda ; Eric Darve
HIGHLIGHT:      In this paper, we propose an approach for embedding imputation which uses grounded information in the form of a knowledge graph.

57, TITLE:      Massive Choice, Ample Tasks (MaChAmp):A Toolkit for Multi-task Learning in NLP
http://arxiv.org/abs/2005.14672
AUTHORS:        Rob van der Goot ; Ahmet Üstün ; Alan Ramponi ; Barbara Plank
COMMENTS:       https://machamp-nlp.github.io/
HIGHLIGHT:      In this paper we present MaChAmp, a toolkit for easy use of fine-tuning BERT-like models in multi-task settings.

58, TITLE:      K-TanH: Efficient TanH For Deep Learning
http://arxiv.org/abs/1909.07729
AUTHORS:        Abhisek Kundu ; Alex Heinecke ; Dhiraj Kalamkar ; Sudarshan Srinivasan ; Eric C. Qin ; Naveen K. Mellempudi ; Dipankar Das ; Kunal Banerjee ; Bharat Kaul ; Pradeep Dubey
COMMENTS:       6 pages, 1 figures
HIGHLIGHT:      We propose K-TanH, a novel, highly accurate, hardware efficient approximation of popular activation function TanH for Deep Learning.

59, TITLE:      HRINet: Alternative Supervision Network for High-resolution CT image Interpolation
http://arxiv.org/abs/2002.04455
AUTHORS:        Jiawei Li ; Jae Chul Koh ; Won-Sook Lee
HIGHLIGHT:      We propose a novel network, High Resolution Interpolation Network (HRINet), aiming at producing high-resolution CT image interpolations.

60, TITLE:      Enabling Incremental Knowledge Transfer for Object Detection at the Edge
http://arxiv.org/abs/2004.05746
AUTHORS:        Mohammad Farhadi Bajestani ; Mehdi Ghasemi ; Sarma Vrudhula ; Yezhou Yang
COMMENTS:       2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (CVPRW)
HIGHLIGHT:      In this paper, a system-level design is proposed to improve the energy consumption of object detection on the user-end device.

61, TITLE:      A Spiking Neural Network Emulating the Structure of the Oculomotor System Requires No Learning to Control a Biomimetic Robotic Head
http://arxiv.org/abs/2002.07534
AUTHORS:        Praveenram Balachandar ; Konstantinos P. Michmizos
COMMENTS:       6 pages, 3 figures, IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob) 2020
HIGHLIGHT:      Leveraging on recent studies on the neural connectome associated with eye movements, we designed a neuromorphic oculomotor controller and placed it at the heart of our in-house biomimetic robotic head prototype.

62, TITLE:      A Safety Framework for Critical Systems Utilising Deep Neural Networks
http://arxiv.org/abs/2003.05311
AUTHORS:        Xingyu Zhao ; Alec Banks ; James Sharp ; Valentin Robu ; David Flynn ; Michael Fisher ; Xiaowei Huang
COMMENTS:       Accepted by SafeComp2020
HIGHLIGHT:      Working towards addressing this challenge, this paper presents a principled novel safety argument framework for critical systems that utilise deep neural networks.

63, TITLE:      Exploration Based Language Learning for Text-Based Games
http://arxiv.org/abs/2001.08868
AUTHORS:        Andrea Madotto ; Mahdi Namazifar ; Joost Huizinga ; Piero Molino ; Adrien Ecoffet ; Huaixiu Zheng ; Alexandros Papangelis ; Dian Yu ; Chandra Khatri ; Gokhan Tur
COMMENTS:       Accepted at IJCAI 2020
HIGHLIGHT:      In this work, we propose to use the exploration approach of Go-Explore for solving text-based games.

64, TITLE:      Egocentric Human Segmentation for Mixed Reality
http://arxiv.org/abs/2005.12074
AUTHORS:        Andrija Gajic ; Ester Gonzalez-Sosa ; Diego Gonzalez-Morin ; Marcos Escudero-Viñolo ; Alvaro Villegas
COMMENTS:       Accepted for presentation at EPIC@CVPR2020 workshop
HIGHLIGHT:      The objective of this work is to segment human body parts from egocentric video using semantic segmentation networks. Our contribution is two-fold: i) we create a semi-synthetic dataset composed of more than 15, 000 realistic images and associated pixel-wise labels of egocentric human body parts, such as arms or legs including different demographic factors; ii) building upon the ThunderNet architecture, we implement a deep learning semantic segmentation algorithm that is able to perform beyond real-time requirements (16 ms for 720 x 720 images).

65, TITLE:      Structured Domain Adaptation with Online Relation Regularization for Unsupervised Person Re-ID
http://arxiv.org/abs/2003.06650
AUTHORS:        Yixiao Ge ; Feng Zhu ; Rui Zhao ; Hongsheng Li
HIGHLIGHT:      To tackle the challenge, we propose an end-to-end structured domain adaptation framework with an online relation-consistency regularization term.

66, TITLE:      When Do Drivers Concentrate? Attention-based Driver Behavior Modeling With Deep Reinforcement Learning
http://arxiv.org/abs/2002.11385
AUTHORS:        Xingbo Fu ; Feng Gao ; Jiang Wu
HIGHLIGHT:      In this paper, we propose an actor-critic method - Attention-based Twin Delayed Deep Deterministic policy gradient (ATD3) algorithm to approximate a driver' s action according to observations and measure the driver' s attention allocation for consecutive time steps in car-following model.

67, TITLE:      Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels
http://arxiv.org/abs/2004.13649
AUTHORS:        Ilya Kostrikov ; Denis Yarats ; Rob Fergus
HIGHLIGHT:      We propose a simple data augmentation technique that can be applied to standard model-free reinforcement learning algorithms, enabling robust learning directly from pixels without the need for auxiliary losses or pre-training.

68, TITLE:      Encoding Visual Attributes in Capsules for Explainable Medical Diagnoses
http://arxiv.org/abs/1909.05926
AUTHORS:        Rodney LaLonde ; Drew Torigian ; Ulas Bagci
COMMENTS:       Accepted for publication at MICCAI 2020 (23rd International Conference on Medical Image Computing and Computer Assisted Intervention). Code is publicly available at https://github.com/lalonderodney/X-Caps
HIGHLIGHT:      We demonstrate a simple 2D capsule network can outperform a state-of-the-art deep dense dual-path 3D CNN at capturing visually-interpretable high-level attributes and malignancy prediction, while providing malignancy prediction scores approaching that of non-explainable 3D CNNs.

69, TITLE:      Noise Estimation Using Density Estimation for Self-Supervised Multimodal Learning
http://arxiv.org/abs/2003.03186
AUTHORS:        Elad Amrani ; Rami Ben-Ari ; Daniel Rotman ; Alex Bronstein
COMMENTS:       Added theoretical analysis and qualitative results
HIGHLIGHT:      In this work, we show that the problem of noise estimation for multimodal data can be reduced to a multimodal density estimation task.

70, TITLE:      Almost Optimal Distribution-free Junta Testing
http://arxiv.org/abs/1901.00717
AUTHORS:        Nader H. Bshouty
COMMENTS:       arXiv admin note: text overlap with arXiv:1802.04859 by other authors
HIGHLIGHT:      In this paper, we give a simple two-sided error adaptive algorithm that makes $\tilde O(k/\epsilon)$ queries.

71, TITLE:      Mixture of Inference Networks for VAE-based Audio-visual Speech Enhancement
http://arxiv.org/abs/1912.10647
AUTHORS:        Mostafa Sadeghi ; Xavier Alameda-Pineda
HIGHLIGHT:      In this paper, we are interested in unsupervised (unknown noise) speech enhancement using latent variable generative models.

72, TITLE:      Self-supervised learning for autonomous vehicles perception: A conciliation between analytical and learning methods
http://arxiv.org/abs/1910.01636
AUTHORS:        Florent Chiaroni ; Mohamed-Cherif Rahal ; Nicolas Hueber ; Frederic Dufaux
HIGHLIGHT:      In this tutorial-style article, we present an overview and a general formalization of the concept of self-supervised learning (SSL) for autonomous vehicles perception.

73, TITLE:      PatchGuard: Provable Defense against Adversarial Patches Using Masks on Small Receptive Fields
http://arxiv.org/abs/2005.10884
AUTHORS:        Chong Xiang ; Arjun Nitin Bhagoji ; Vikash Sehwag ; Prateek Mittal
HIGHLIGHT:      In this paper, we propose a general defense framework called PatchGuard that can achieve both high clean accuracy and provable robustness against localized adversarial patches.

74, TITLE:      An Epistemic Approach to the Formal Specification of Statistical Machine Learning
http://arxiv.org/abs/2004.12734
AUTHORS:        Yusuke Kawamoto
COMMENTS:       under submission (the journal version of the SEFM'19 conference paper arxiv:1907.10327)
HIGHLIGHT:      We propose an epistemic approach to formalizing statistical properties of machine learning.

75, TITLE:      Graph-Based Parallel Large Scale Structure from Motion
http://arxiv.org/abs/1912.10659
AUTHORS:        Yu Chen ; Shuhan Shen ; Yisong Chen ; Guoping Wang
COMMENTS:       In submission to Pattern Recognition 2020
HIGHLIGHT:      In this work, large scale SfM is deemed as a graph problem, and we tackle it in a divide-and-conquer manner.

76, TITLE:      Learning Shared Cross-modality Representation Using Multispectral-LiDAR and Hyperspectral Data
http://arxiv.org/abs/1912.08837
AUTHORS:        Danfeng Hong ; Jocelyn Chanussot ; Naoto Yokoya ; Jian Kang ; Xiao Xiang Zhu
HIGHLIGHT:      To this end, in this letter, we propose to learn a shared feature space across multi-modalities in the training process.

77, TITLE:      Policy-Driven Neural Response Generation for Knowledge-Grounded Dialogue Systems
http://arxiv.org/abs/2005.12529
AUTHORS:        Behnam Hedayatnia ; Seokhwan Kim ; Yang Liu ; Karthik Gopalakrishnan ; Mihail Eric ; Dilek Hakkani-Tur
COMMENTS:       Typos in Figure 2 and 6
HIGHLIGHT:      In this paper, we propose using a dialogue policy to plan the content and style of target responses in the form of an action plan, which includes knowledge sentences related to the dialogue context, targeted dialogue acts, topic information, etc.

78, TITLE:      Adaptive Loss-aware Quantization for Multi-bit Networks
http://arxiv.org/abs/1912.08883
AUTHORS:        Zhongnan Qu ; Zimu Zhou ; Yun Cheng ; Lothar Thiele
COMMENTS:       To appear in CVPR 2020
HIGHLIGHT:      We propose Adaptive Loss-aware Quantization (ALQ), a new MBN quantization pipeline that is able to achieve an average bitwidth below one-bit without notable loss in inference accuracy.

79, TITLE:      Deep Multi-Agent Reinforcement Learning for Decentralized Continuous Cooperative Control
http://arxiv.org/abs/2003.06709
AUTHORS:        Christian Schroeder de Witt ; Bei Peng ; Pierre-Alexandre Kamienny ; Philip Torr ; Wendelin Böhmer ; Shimon Whiteson
HIGHLIGHT:      To make progress in CTDE, we introduce Multi-Agent Mujoco, a novel benchmark suite that, unlike StarCraft II, the predominant benchmark environment, applies to continuous robotic control tasks.

80, TITLE:      Derivative Manipulation for General Example Weighting
http://arxiv.org/abs/1905.11233
AUTHORS:        Xinshao Wang ; Elyor Kodirov ; Yang Hua ; Neil M. Robertson
COMMENTS:       Is it the right time to make a change from the design of loss function to the design of derivative directly?
HIGHLIGHT:      Therefore, we propose derivative manipulation (DM), a novel and general example weighting approach for training robust deep models under these adverse conditions.

81, TITLE:      Sequence Information Channel Concatenation for Improving Camera Trap Image Burst Classification
http://arxiv.org/abs/2005.00116
AUTHORS:        Bhuvan Malladihalli Shashidhara ; Darshan Mehta ; Yash Kale ; Dan Morris ; Megan Hazen
COMMENTS:       8 pages, 4 figures, 2 tables. Git repository can be found at: https://github.com/bhuvi3/camera_trap_animal_classification
HIGHLIGHT:      In this work, we explore a variety of approaches, and measure the impact of using short image sequences (burst of 3 images) on improving the camera trap image classification.

82, TITLE:      Deep-URL: A Model-Aware Approach To Blind Deconvolution Based On Deep Unfolded Richardson-Lucy Network
http://arxiv.org/abs/2002.01053
AUTHORS:        Chirag Agarwal ; Shahin Khobahi ; Arindam Bose ; Mojtaba Soltanalian ; Dan Schonfeld
COMMENTS:       Accepted. 27th IEEE International Conference on Image Processing (ICIP), 2020
HIGHLIGHT:      In this paper, we consider the problem of blind deconvolution and propose a novel model-aware deep architecture that allows for the recovery of both the blur kernel and the sharp image from the blurred image.

83, TITLE:      GAN Compression: Efficient Architectures for Interactive Conditional GANs
http://arxiv.org/abs/2003.08936
AUTHORS:        Muyang Li ; Ji Lin ; Yaoyao Ding ; Zhijian Liu ; Jun-Yan Zhu ; Song Han
HIGHLIGHT:      In this work, we propose a general-purpose compression framework for reducing the inference time and model size of the generator in cGANs.

84, TITLE:      Recognizing Families In the Wild: White Paper for the 4th Edition Data Challenge
http://arxiv.org/abs/2002.06303
AUTHORS:        Joseph P. Robinson ; Yu Yin ; Zaid Khan ; Ming Shao ; Siyu Xia ; Michael Stopa ; Samson Timoner ; Matthew A. Turk ; Rama Chellappa ; Yun Fu
COMMENTS:       White Paper for challenge in conjunction with 15th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2020)
HIGHLIGHT:      In the end, the purpose of this paper is to describe the 2020 RFIW challenge, end-to-end, along with forecasts in promising future directions.

85, TITLE:      On subset sum problem in branch groups
http://arxiv.org/abs/2006.03470
AUTHORS:        Andrey Nikolaev ; Alexander Ushakov
COMMENTS:       v2: changed \documentclass. 6 pages
HIGHLIGHT:      In this brief note, we show that the subset sum problem is NP-complete in the first Grigorchuk group.

86, TITLE:      On the Limitations of Cross-lingual Encoders as Exposed by Reference-Free Machine Translation Evaluation
http://arxiv.org/abs/2005.01196
AUTHORS:        Wei Zhao ; Goran Glavaš ; Maxime Peyrard ; Yang Gao ; Robert West ; Steffen Eger
COMMENTS:       ACL2020 Camera Ready (v3: several small fixes, e.g., Unicode errors)
HIGHLIGHT:      In this paper, we concern ourselves with reference-free machine translation (MT) evaluation where we directly compare source texts to (sometimes low-quality) system translations, which represents a natural adversarial setup for multilingual encoders.

87, TITLE:      Inherent Tradeoffs in Learning Fair Representations
http://arxiv.org/abs/1906.08386
AUTHORS:        Han Zhao ; Geoffrey J. Gordon
COMMENTS:       A tighter bound on accuracy parity in Theorem 3.3
HIGHLIGHT:      In this paper, through the lens of information theory, we provide the first result that quantitatively characterizes the tradeoff between demographic parity and the joint utility across different population groups.

88, TITLE:      Fast Complete Algorithm for Multiplayer Nash Equilibrium
http://arxiv.org/abs/2002.04734
AUTHORS:        Sam Ganzfried
HIGHLIGHT:      We describe a new complete algorithm for computing Nash equilibrium in multiplayer general-sum games, based on a quadratically-constrained feasibility program formulation.

89, TITLE:      Model-guided Multi-path Knowledge Aggregation for Aerial Saliency Prediction
http://arxiv.org/abs/1811.05625
AUTHORS:        Kui Fu ; Jia Li ; Yu Zhang ; Hongze Shen ; Yonghong Tian
HIGHLIGHT:      Based on this dataset, we propose a Model-guided Multi-path Network (MM-Net) that serves as a baseline model for aerial video saliency prediction. To investigate these challenges, this paper proposes a large-scale video dataset for aerial saliency prediction, which consists of ground-truth salient object regions of 1,000 aerial videos, annotated by 24 subjects.

90, TITLE:      Optimal Farsighted Agents Tend to Seek Power
http://arxiv.org/abs/1912.01683
AUTHORS:        Alexander Matt Turner ; Logan Smith ; Rohin Shah ; Prasad Tadepalli
COMMENTS:       9 pages, 25 pages with appendices
HIGHLIGHT:      Others have voiced skepticism: humans seem idiosyncratic in their urges to power, which need not be present in the agents we design.

91, TITLE:      An Analysis of the Utility of Explicit Negative Examples to Improve the Syntactic Abilities of Neural Language Models
http://arxiv.org/abs/2004.02451
AUTHORS:        Hiroshi Noji ; Hiroya Takamura
COMMENTS:       ACL 2020 camera ready (long paper); code is available at https://github.com/aistairc/lm_syntax_negative
HIGHLIGHT:      We explore the utilities of explicit negative examples in training neural language models.

92, TITLE:      LiSSS: A toy corpus of Spanish Literary Sentences for Emotions detection
http://arxiv.org/abs/2005.08223
AUTHORS:        Juan-Manuel Torres-Moreno ; Luis-Gil Moreno-Jiménez
COMMENTS:       8 pages, 3 tables
HIGHLIGHT:      In this work we present a new small data-set in Computational Creativity (CC) field, the Spanish Literary Sentences for emotions detection corpus (LISSS).

93, TITLE:      SDCT-AuxNet$^θ$: DCT Augmented Stain Deconvolutional CNN with Auxiliary Classifier for Cancer Diagnosis
http://arxiv.org/abs/2006.00304
AUTHORS:        Shiv Gehlot ; Anubha Gupta ; Ritu Gupta
COMMENTS:       The final version of this preprint has been published in Medical Image Analysis
HIGHLIGHT:      This paper discusses the recent release of a large dataset and presents a novel deep learning architecture for the classification of cell images of ALL cancer.

94, TITLE:      Exploiting Temporal Coherence for Multi-modal Video Categorization
http://arxiv.org/abs/2002.03844
AUTHORS:        Palash Goyal ; Saurabh Sahu ; Shalini Ghosh ; Chul Lee
HIGHLIGHT:      In this paper, we focus on the problem of video categorization by using a multimodal approach.

95, TITLE:      Making Logic Learnable With Neural Networks
http://arxiv.org/abs/2002.03847
AUTHORS:        Tobias Brudermueller ; Dennis L. Shung ; Adrian J. Stanley ; Johannes Stegmaier ; Smita Krishnaswamy
HIGHLIGHT:      We propose a novel logic learning pipeline that combines the advantages of neural networks and logic circuits.

96, TITLE:      Deep Predictive Motion Tracking in Magnetic Resonance Imaging: Application to Fetal Imaging
http://arxiv.org/abs/1909.11625
AUTHORS:        Ayush Singh ; Seyed Sadegh Mohseni Salehi ; Ali Gholipour
COMMENTS:       The article has been published in IEEE TMI: 14 pages, 11 figures, 2 tables and 1 supplementary https://github.com/bchimagine/DeepPredictiveMotionTracking
HIGHLIGHT:      We compared our trained network on held out test sets (including data with different characteristics, e.g. different fetuses scanned at different ages, and motion trajectories recorded from volunteer subjects) with networks designed for estimation as well as methods adopted to make predictions.

97, TITLE:      Breaking the cycle -- Colleagues are all you need
http://arxiv.org/abs/1911.10538
AUTHORS:        Ori Nizan ; Ayellet Tal
HIGHLIGHT:      This paper proposes a novel approach to performing image-to-image translation between unpaired domains.

98, TITLE:      Semi-Supervised Semantic Segmentation with Cross-Consistency Training
http://arxiv.org/abs/2003.09005
AUTHORS:        Yassine Ouali ; Céline Hudelot ; Myriam Tami
COMMENTS:       Published at CVPR 2020
HIGHLIGHT:      In this paper, we present a novel cross-consistency based semi-supervised approach for semantic segmentation.

99, TITLE:      Architectural Resilience to Foreground-and-Background Adversarial Noise
http://arxiv.org/abs/2003.10045
AUTHORS:        Carl Cheng ; Evan Hu
COMMENTS:       9 pages, 8 figures; updated email addresses
HIGHLIGHT:      In this work, we instead propose distinct model-agnostic benchmark perturbations of images in order to investigate the resilience and robustness of different network architectures.

100, TITLE:     Proximity-based Networking: Small world overlays optimized with particle swarm optimization
http://arxiv.org/abs/2006.02006
AUTHORS:        Chase Smith ; Alex Rusnak
HIGHLIGHT:      We propose a networking scheme that incorporates geographic location in chord for the organization of peers within each node's partitioned key space.

101, TITLE:     FA-GANs: Facial Attractiveness Enhancement with Generative Adversarial Networks on Frontal Faces
http://arxiv.org/abs/2005.08168
AUTHORS:        Jingwu He ; Chuan Wang ; Yang Zhang ; Jie Guo ; Yanwen Guo
COMMENTS:       10 pages, 8 figures
HIGHLIGHT:      In this paper, we propose the first Generative Adversarial Networks (GANs) for enhancing facial attractiveness in both geometry and appearance aspects, which we call "FA-GANs".

102, TITLE:     Self-Paced Deep Reinforcement Learning
http://arxiv.org/abs/2004.11812
AUTHORS:        Pascal Klink ; Carlo D'Eramo ; Jan Peters ; Joni Pajarinen
HIGHLIGHT:      In this paper, we propose an answer by interpreting the curriculum generation as an inference problem, where distributions over tasks are progressively learned to approach the target task.

103, TITLE:     Data augmentation with Mobius transformations
http://arxiv.org/abs/2002.02917
AUTHORS:        Sharon Zhou ; Jiequan Zhang ; Hang Jiang ; Torbjorn Lundh ; Andrew Y. Ng
HIGHLIGHT:      In this paper, we present a novel method of applying Mobius transformations to augment input images during training.

104, TITLE:     GATCluster: Self-Supervised Gaussian-Attention Network for Image Clustering
http://arxiv.org/abs/2002.11863
AUTHORS:        Chuang Niu ; Jun Zhang ; Ge Wang ; Jimin Liang
HIGHLIGHT:      We propose a self-supervised Gaussian ATtention network for image Clustering (GATCluster).

105, TITLE:     Demographics Should Not Be the Reason of Toxicity: Mitigating Discrimination in Text Classifications with Instance Weighting
http://arxiv.org/abs/2004.14088
AUTHORS:        Guanhua Zhang ; Bing Bai ; Junqi Zhang ; Kun Bai ; Conghui Zhu ; Tiejun Zhao
COMMENTS:       Accepted by ACL 2020
HIGHLIGHT:      In this paper, we formalize the unintended biases in text classification datasets as a kind of selection bias from the non-discrimination distribution to the discrimination distribution.

106, TITLE:     Antenna Optimization Using a New Evolutionary Algorithm Based on Tukey-Lambda Probability Distribution
http://arxiv.org/abs/2005.13594
AUTHORS:        Vahraz Jamnejad ; Ahmad Hoorfar
COMMENTS:       5 pages, to be submitted to IEEE ACCESS
HIGHLIGHT:      In this paper, we introduce a new evolutionary optimization algorithm based on Tukey's symmetric lambda distribution.

107, TITLE:     ProSelfLC: Progressive Self Label Correction for Training Robust Deep Neural Networks
http://arxiv.org/abs/2005.03788
AUTHORS:        Xinshao Wang ; Yang Hua ; Elyor Kodirov ; Neil M. Robertson
COMMENTS:       Open discussion: we show it's fine for a learner (student) to be confident towards a correct low entropy status. Then more research attention should be paid to the definition of correct knowledge, as in genera we accept, human annotations used for learning supervision may be biased, subjective, and wrong
HIGHLIGHT:      To resolve these issues, taking two well-accepted expertise: deep neural networks learn meaningful patterns before fitting noise [1] and minimum entropy regularisation principle [2], we propose a simple end-to-end method named ProSelfLC, which is endorsed by long learning time and high prediction confidence.

108, TITLE:     FBNetV3: Joint Architecture-Recipe Search using Neural Acquisition Function
http://arxiv.org/abs/2006.02049
AUTHORS:        Xiaoliang Dai ; Alvin Wan ; Peizhao Zhang ; Bichen Wu ; Zijian He ; Zhen Wei ; Kan Chen ; Yuandong Tian ; Matthew Yu ; Peter Vajda ; Joseph E. Gonzalez
HIGHLIGHT:      To address this oversight, we present JointNAS to search both (a) architectures and (b) their corresponding training recipes.

109, TITLE:     Boosting Adversarial Training with Hypersphere Embedding
http://arxiv.org/abs/2002.08619
AUTHORS:        Tianyu Pang ; Xiao Yang ; Yinpeng Dong ; Kun Xu ; Hang Su ; Jun Zhu
HIGHLIGHT:      In this work, we advocate incorporating the hypersphere embedding (HE) mechanism into the AT procedure by regularizing the features onto compact manifolds, which constitutes a lightweight yet effective module to blend in the strength of representation learning.