2020.06.25.txt

==========New Papers==========
1, TITLE:       Road surface detection and differentiation considering surface damages
http://arxiv.org/abs/2006.13377
AUTHORS:        Thiago Rateke ; Aldo von Wangenheim
COMMENTS:       13 pages
HIGHLIGHT:      Road surface detection and differentiation considering surface damages

2, TITLE:       Deep Generative Model-based Quality Control for Cardiac MRI Segmentation
http://arxiv.org/abs/2006.13379
AUTHORS:        Shuo Wang ; Giacomo Tarroni ; Chen Qin ; Yuanhan Mo ; Chengliang Dai ; Chen Chen ; Ben Glocker ; Yike Guo ; Daniel Rueckert ; Wenjia Bai
COMMENTS:       The paper is accepted to MICCAI 2020
HIGHLIGHT:      Here we propose a novel deep generative model-based framework for quality control of cardiac MRI segmentation.

3, TITLE:       Affinity Fusion Graph-based Framework for Natural Image Segmentation
http://arxiv.org/abs/2006.13542
AUTHORS:        Yang Zhang ; Moyun Liu ; Jingwu He ; Fei Pan ; Yanwen Guo
COMMENTS:       11 pages, 10 figures
HIGHLIGHT:      This paper proposes an affinity fusion graph framework to effectively connect different graphs with highly discriminating power and nonlinearity for natural image segmentation.

4, TITLE:       Crossmodal Language Grounding in an Embodied Neurocognitive Model
http://arxiv.org/abs/2006.13546
AUTHORS:        Stefan Heinrich ; Yuan Yao ; Tobias Hinz ; Zhiyuan Liu ; Thomas Hummel ; Matthias Kerzel ; Cornelius Weber ; Stefan Wermter
COMMENTS:       Under review, 25 pages
HIGHLIGHT:      In this paper, we present a neurocognitive model for language grounding which reflects bio-inspired mechanisms such as an implicit adaptation of timescales as well as end-to-end multimodal abstraction.

5, TITLE:       Post-DAE: Anatomically Plausible Segmentation via Post-Processing with Denoising Autoencoders
http://arxiv.org/abs/2006.13791
AUTHORS:        Agostina J Larrazabal ; César Martínez ; Ben Glocker ; Enzo Ferrante
COMMENTS:       Accepted for publication in IEEE Transactions on Medical Imaging (IEEE TMI)
HIGHLIGHT:      We introduce Post-DAE, a post-processing method based on denoising autoencoders (DAE) to improve the anatomical plausibility of arbitrary biomedical image segmentation algorithms.

6, TITLE:       NASTransfer: Analyzing Architecture Transferability in Large Scale Neural Architecture Search
http://arxiv.org/abs/2006.13314
AUTHORS:        Rameswar Panda ; Michele Merler ; Mayoore Jaiswal ; Hui Wu ; Kandan Ramakrishnan ; Ulrich Finkler ; Chun-Fu Chen ; Minsik Cho ; David Kung ; Rogerio Feris ; Bishwaranjan Bhattacharjee
COMMENTS:       19 pages, 19 Figures, 6 Tables
HIGHLIGHT:      In this work, we propose to analyze the architecture transferability of different NAS methods by performing a series of experiments on large scale benchmarks such as ImageNet1K and ImageNet22K.

7, TITLE:       A Methodology for Creating AI FactSheets
http://arxiv.org/abs/2006.13796
AUTHORS:        John Richards ; David Piorkowski ; Michael Hind ; Stephanie Houde ; Aleksandra Mojsilović
COMMENTS:       18 pages
HIGHLIGHT:      This paper describes this methodology and shares the insights we have gathered.

8, TITLE:       Normalized Loss Functions for Deep Learning with Noisy Labels
http://arxiv.org/abs/2006.13554
AUTHORS:        Xingjun Ma ; Hanxun Huang ; Yisen Wang ; Simone Romano ; Sarah Erfani ; James Bailey
COMMENTS:       Accepted to ICML 2020
HIGHLIGHT:      In this paper, we theoretically show by applying a simple normalization that: any loss can be made robust to noisy labels.

9, TITLE:       ReLoC Reloaded: A Mechanized Relational Logic for Fine-Grained Concurrency and Logical Atomicity
http://arxiv.org/abs/2006.13635
AUTHORS:        Dan Frumin ; Robbert Krebbers ; Lars Birkedal
HIGHLIGHT:      We present a new version of ReLoC: a relational logic for proving refinements of programs in a language with higher-order state, fine-grained concurrency, polymorphism and recursive types.

10, TITLE:      Efficient Constituency Parsing by Pointing
http://arxiv.org/abs/2006.13557
AUTHORS:        Thanh-Tung Nguyen ; Xuan-Phi Nguyen ; Shafiq Joty ; Xiaoli Li
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      We propose a novel constituency parsing model that casts the parsing problem into a series of pointing tasks.

11, TITLE:      Auto-PyTorch Tabular: Multi-Fidelity MetaLearning for Efficient and Robust AutoDL
http://arxiv.org/abs/2006.13799
AUTHORS:        Lucas Zimmer ; Marius Lindauer ; Frank Hutter
HIGHLIGHT:      In this paper, we introduce Auto-PyTorch, which brings the best of these two worlds together by jointly and robustly optimizing the architecture of networks and the training hyperparameters to enable fully automated deep learning (AutoDL). To thoroughly study our assumptions on how to design such an AutoDL system, we additionally introduce a new benchmark on learning curves for DNNs, dubbed LCBench, and run extensive ablation studies of the full Auto-PyTorch on typical AutoML benchmarks, eventually showing that Auto-PyTorch performs better than several state-of-the-art competitors on average.

12, TITLE:      Automatic Estimation of Self-Reported Pain by Interpretable Representations of Motion Dynamics
http://arxiv.org/abs/2006.13882
AUTHORS:        Benjamin Szczapa ; Mohamed Daoudi ; Stefano Berretti ; Pietro Pala ; Alberto Del Bimbo ; Zakia Hammal
COMMENTS:       accepted at ICPR 2020 Conference
HIGHLIGHT:      We propose an automatic method for pain intensity measurement from video.

13, TITLE:      Noetherian operators and primary decomposition
http://arxiv.org/abs/2006.13881
AUTHORS:        Justin Chen ; Marc Härkönen ; Robert Krone ; Anton Leykin
COMMENTS:       17 pages, codebase available at https://github.com/haerski/NoetherianOperators
HIGHLIGHT:      We develop a framework, as well as algorithms, for computing Noetherian operators with local dual spaces, both symbolically and numerically.

14, TITLE:      Differentiable Window for Dynamic Local Attention
http://arxiv.org/abs/2006.13561
AUTHORS:        Thanh-Tung Nguyen ; Xuan-Phi Nguyen ; Shafiq Joty ; Xiaoli Li
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      We propose Differentiable Window, a new neural module and general purpose component for dynamic window selection.

15, TITLE:      Learning for Video Compression with Recurrent Auto-Encoder and Recurrent Probability Model
http://arxiv.org/abs/2006.13560
AUTHORS:        Ren Yang ; Fabian Mentzer ; Luc Van Gool ; Radu Timofte
COMMENTS:       This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
HIGHLIGHT:      To overcome this shortcoming, this paper proposes a Recurrent Learned Video Compression (RLVC) approach with the Recurrent Auto-Encoder (RAE) and Recurrent Probability Model (RPM).

16, TITLE:      DISK: Learning local features with policy gradient
http://arxiv.org/abs/2006.13566
AUTHORS:        Michał J. Tyszkiewicz ; Pascal Fua ; Eduard Trulls
HIGHLIGHT:      We introduce DISK (DIScrete Keypoints), a novel method that overcomes these obstacles by leveraging principles from Reinforcement Learning (RL), optimizing end-to-end for a high number of correct feature matches.

17, TITLE:      Realistic Adversarial Data Augmentation for MR Image Segmentation
http://arxiv.org/abs/2006.13322
AUTHORS:        Chen Chen ; Chen Qin ; Huaqi Qiu ; Cheng Ouyang ; Shuo Wang ; Liang Chen ; Giacomo Tarroni ; Wenjia Bai ; Daniel Rueckert
COMMENTS:       13 pages. This paper is accepted to MICCAI 2020
HIGHLIGHT:      In this work, we propose an adversarial data augmentation method for training neural networks for medical image segmentation.

18, TITLE:      Classifying Referential and Non-referential It Using Gaze
http://arxiv.org/abs/2006.13327
AUTHORS:        Victoria Yaneva ; Le An Ha ; Richard Evans ; Ruslan Mitkov
HIGHLIGHT:      In this paper, we use eye-tracking data to learn how humans perform this disambiguation.

19, TITLE:      Safe Learning under Uncertain Objectives and Constraints
http://arxiv.org/abs/2006.13326
AUTHORS:        Mohammad Fereydounian ; Zebang Shen ; Aryan Mokhtari ; Amin Karbasi ; Hamed Hassani
COMMENTS:       42 pages, 2 figures
HIGHLIGHT:      In this paper, we consider non-convex optimization problems under \textit{unknown} yet safety-critical constraints.

20, TITLE:      On the Empirical Neural Tangent Kernel of Standard Finite-Width Convolutional Neural Network Architectures
http://arxiv.org/abs/2006.13645
AUTHORS:        Maxim Samarin ; Volker Roth ; David Belius
COMMENTS:       10 pages
HIGHLIGHT:      The Neural Tangent Kernel (NTK) is an important milestone in the ongoing effort to build a theory for deep learning.

21, TITLE:      Control-Aware Representations for Model-based Reinforcement Learning
http://arxiv.org/abs/2006.13408
AUTHORS:        Brandon Cui ; Yinlam Chow ; Mohammad Ghavamzadeh
HIGHLIGHT:      In this paper, we take a few steps towards addressing these questions.

22, TITLE:      AutoKnow: Self-Driving Knowledge Collection for Products of Thousands of Types
http://arxiv.org/abs/2006.13473
AUTHORS:        Xin Luna Dong ; Xiang He ; Andrey Kan ; Xian Li ; Yan Liang ; Jun Ma ; Yifan Ethan Xu ; Chenwei Zhang ; Tong Zhao ; Gabriel Blanco Saldana ; Saurabh Deshpande ; Alexandre Michetti Manduca ; Jay Ren ; Surender Pal Singh ; Fan Xiao ; Haw-Shiuan Chang ; Giannis Karamanolakis ; Yuning Mao ; Yaqing Wang ; Christos Faloutsos ; Andrew McCallum ; Jiawei Han
COMMENTS:       KDD 2020
HIGHLIGHT:      We describe AutoKnow, our automatic (self-driving) system that addresses these challenges.

23, TITLE:      Malignancy-Aware Follow-Up Volume Prediction for Lung Nodules
http://arxiv.org/abs/2006.13890
AUTHORS:        Yamin Li ; Jiancheng Yang ; Yi Xu ; Jingwei Xu ; Xiaodan Ye ; Guangyu Tao ; Xueqian Xie ; Guixue Liu
COMMENTS:       MICCAI 2020, with supplementary materials
HIGHLIGHT:      To this end, we propose a unified framework, named Nodule Follow-Up Prediction Network (NoFoNet), which predicts the growth of pulmonary nodules with high-quality visual appearances and accurate quantitative malignancy scores, given any time interval from baseline observations. We build an in-house follow-up dataset from two medical centers to validate the effectiveness of the proposed method.

24, TITLE:      Modelling the Statistics of Cyclic Activities by Trajectory Analysis on the Manifold of Positive-Semi-Definite Matrices
http://arxiv.org/abs/2006.13895
AUTHORS:        Ettore Maria Celozzi ; Luca Ciabini ; Luca Cultrera ; Pietro Pala ; Stefano Berretti ; Mohamed Daoudi ; Alberto Del Bimbo
COMMENTS:       accepted at 15th IEEE International Conference on Automatic Face and Gesture Recognition 2020
HIGHLIGHT:      In this paper, a model is presented to extract statistical summaries to characterize the repetition of a cyclic body action, for instance a gym exercise, for the purpose of checking the compliance of the observed action to a template one and highlighting the parts of the action that are not correctly executed (if any).

25, TITLE:      Exploring the Security Awareness of the Python and JavaScript Open Source Communities
http://arxiv.org/abs/2006.13652
AUTHORS:        Gábor Antal ; Márton Keleti ; Péter Hegedűs
COMMENTS:       17th International Conference on Mining Software Repositories
HIGHLIGHT:      By analyzing large quantities of commits in the open-source communities, we can categorize the vulnerabilities mitigated by the developers and study their distribution, resolution time, etc. to learn and improve security management processes and practices.

26, TITLE:      Neural Non-Rigid Tracking
http://arxiv.org/abs/2006.13240
AUTHORS:        Aljaž Božič ; Pablo Palafox ; Michael Zollhöfer ; Angela Dai ; Justus Thies ; Matthias Nießner
COMMENTS:       Video: https://youtu.be/nqYaxM6Rj8I
HIGHLIGHT:      We introduce a novel, end-to-end learnable, differentiable non-rigid tracker that enables state-of-the-art non-rigid reconstruction.

27, TITLE:      Accelerated Large Batch Optimization of BERT Pretraining in 54 minutes
http://arxiv.org/abs/2006.13484
AUTHORS:        Shuai Zheng ; Haibin Lin ; Sheng Zha ; Mu Li
COMMENTS:       Technical Report
HIGHLIGHT:      In this paper, we propose an accelerated gradient method called LANS to improve the efficiency of using large mini-batches for training.

28, TITLE:      Labelling unlabelled videos from scratch with multi-modal self-supervision
http://arxiv.org/abs/2006.13662
AUTHORS:        Yuki M. Asano ; Mandela Patrick ; Christian Rupprecht ; Andrea Vedaldi
COMMENTS:       project page: https://www.robots.ox.ac.uk/~vgg/research/selavi
HIGHLIGHT:      In this work, we a) show that unsupervised labelling of a video dataset does not come for free from strong feature encoders and b) propose a novel clustering method that allows pseudo-labelling of a video dataset without any human annotations, by leveraging the natural correspondence between the audio and visual modalities.

29, TITLE:      Automating Text Naturalness Evaluation of NLG Systems
http://arxiv.org/abs/2006.13268
AUTHORS:        Erion Çano ; Ondřej Bojar
COMMENTS:       15 pages, 4 equations, 3 tables. arXiv admin note: text overlap with arXiv:2006.03189
HIGHLIGHT:      We present here an attempt to automate the evaluation of text naturalness which is a very important characteristic of natural language generation methods.

30, TITLE:      Learning Interclass Relations for Image Classification
http://arxiv.org/abs/2006.13491
AUTHORS:        Muhamedrahimov Raouf ; Bar Amir ; Akselrod-Ballin Ayelet
HIGHLIGHT:      In this work, we propose novel formulations of the classification problem, based on a realization that the assumption of class-independence is a limiting factor that leads to the requirement of more training data.

31, TITLE:      Robot Object Retrieval with Contextual Natural Language Queries
http://arxiv.org/abs/2006.13253
AUTHORS:        Thao Nguyen ; Nakul Gopalan ; Roma Patel ; Matt Corsaro ; Ellie Pavlick ; Stefanie Tellex
HIGHLIGHT:      We develop a model to retrieve objects based on descriptions of their usage. We also present a new dataset of 655 verb-object pairs denoting object usage over 50 verbs and 216 object classes.

32, TITLE:      Iris Presentation Attack Detection: Where Are We Now?
http://arxiv.org/abs/2006.13252
AUTHORS:        Aidan Boyd ; Zhaoyuan Fang ; Adam Czajka ; Kevin W. Bowyer
COMMENTS:       Under revision for Pattern Recognition Letters
HIGHLIGHT:      This work presents an overview of the most important advances in the area of iris presentation attack detection published in recent two years.

33, TITLE:      Adversarial Soft Advantage Fitting: Imitation Learning without Policy Optimization
http://arxiv.org/abs/2006.13258
AUTHORS:        Paul Barde ; Julien Roy ; Wonseok Jeon ; Joelle Pineau ; Christopher Pal ; Derek Nowrouzezahrai
HIGHLIGHT:      We propose to remove the burden of the policy optimization steps by leveraging a novel discriminator formulation.

34, TITLE:      Rescaling Egocentric Vision
http://arxiv.org/abs/2006.13256
AUTHORS:        Dima Damen ; Hazel Doughty ; Giovanni Maria Farinella ; Antonino Furnari ; Evangelos Kazakos ; Jian Ma ; Davide Moltisanti ; Jonathan Munro ; Toby Perrett ; Will Price ; Michael Wray
COMMENTS:       Dataset available from: http://epic-kitchens.github.io/
HIGHLIGHT:      This paper introduces EPIC-KITCHENS-100, the largest annotated egocentric dataset - 100 hrs, 20M frames, 90K actions - of wearable videos capturing long-term unscripted activities in 45 environments.

35, TITLE:      Was there COVID-19 back in 2012? Challenge for AI in Diagnosis with Similar Indications
http://arxiv.org/abs/2006.13262
AUTHORS:        Imon Banerjee ; Priyanshu Sinha ; Saptarshi Purkayastha ; Nazanin Mashhaditafreshi ; Amara Tariq ; Jiwoong Jeong ; Hari Trivedi ; Judy W. Gichoya
HIGHLIGHT:      Methods: In this paper, we present our argument regarding the efficiency and applicability of existing deep learning models for COVID-19 diagnosis.

36, TITLE:      Anomaly Detection with Deep Perceptual Autoencoders
http://arxiv.org/abs/2006.13265
AUTHORS:        Nina Tuluptceva ; Bart Bakker ; Irina Fedulova ; Heinrich Schulz ; Dmitry V. Dylov
COMMENTS:       A preprint
HIGHLIGHT:      To address this problem, we introduce a new powerful method of image anomaly detection.

37, TITLE:      Image-to-image Mapping with Many Domains by Sparse Attribute Transfer
http://arxiv.org/abs/2006.13291
AUTHORS:        Matthew Amodio ; Rim Assouel ; Victor Schmidt ; Tristan Sylvain ; Smita Krishnaswamy ; Yoshua Bengio
HIGHLIGHT:      We propose an alternate approach that directly restricts the generator to performing a simple sparse transformation in a latent layer, motivated by recent work from cognitive neuroscience suggesting an architectural prior on representations corresponding to consciousness.

38, TITLE:      Supervised Understanding of Word Embeddings
http://arxiv.org/abs/2006.13299
AUTHORS:        Halid Ziya Yerebakan ; Parmeet Bhatia ; Yoshihisa Shinagawa
HIGHLIGHT:      In this study, we have obtained supervised projections in the form of the linear keyword-level classifiers on word embeddings.

39, TITLE:      Thalamocortical motor circuit insights for more robust hierarchical control of complex sequences
http://arxiv.org/abs/2006.13332
AUTHORS:        Laureline Logiaco ; G. Sean Escola
COMMENTS:       14 pages, 5 figures. Submitted to NeurIPS 2020
HIGHLIGHT:      We study learning of recurrent neural networks that produce temporal sequences consisting of the concatenation of re-usable "motifs".

40, TITLE:      Large-scale detection and categorization of oil spills from SAR images with deep learning
http://arxiv.org/abs/2006.13575
AUTHORS:        Filippo Maria Bianchi ; Martine M. Espeseth ; Njål Borch
HIGHLIGHT:      We propose a deep learning framework to detect and categorize oil spills in synthetic aperture radar (SAR) images at a large scale.

41, TITLE:      One Model to Pronounce Them All: Multilingual Grapheme-to-Phoneme Conversion With a Transformer Ensemble
http://arxiv.org/abs/2006.13343
AUTHORS:        Kaili Vesik ; Muhammad Abdul-Mageed ; Miikka Silfverberg
COMMENTS:       7 pages, submitted to SIGMORPHON 2020 Shared Task 1
HIGHLIGHT:      We describe a simple approach of exploiting model ensembles, based on multilingual Transformers and self-training, to develop a highly effective G2P solution for 15 languages.

42, TITLE:      Applying Lie Groups Approaches for Rigid Registration of Point Clouds
http://arxiv.org/abs/2006.13341
AUTHORS:        Liliane Rodrigues de Almeida ; Gilson A. Giraldi ; Marcelo Bernardes Vieira
COMMENTS:       29 pages, 4 figures, 1 table
HIGHLIGHT:      In this paper we focus on application of Lie groups and Lie algebras to find the rigid transformation that best register two surfaces represented by point clouds.

43, TITLE:      Principal Component Networks: Parameter Reduction Early in Training
http://arxiv.org/abs/2006.13347
AUTHORS:        Roger Waleffe ; Theodoros Rekatsinas
HIGHLIGHT:      In this paper, we show how to find small networks that exhibit the same performance as their overparameterized counterparts after only a few training epochs.

44, TITLE:      Rethinking Distributional Matching Based Domain Adaptation
http://arxiv.org/abs/2006.13352
AUTHORS:        Bo Li ; Yezhen Wang ; Tong Che ; Shanghang Zhang ; Sicheng Zhao ; Pengfei Xu ; Wei Zhou ; Yoshua Bengio ; Kurt Keutzer
HIGHLIGHT:      We hope our intuitive yet effective method will serve as a useful new direction and increase the robustness of DA in real scenarios. In this paper, in order to devise robust DA algorithms, we first systematically analyze the limitations of DM based methods, and then build new benchmarks with more realistic domain shifts to evaluate the well-accepted DM methods.

45, TITLE:      Retrospective Loss: Looking Back to Improve Training of Deep Neural Networks
http://arxiv.org/abs/2006.13593
AUTHORS:        Surgan Jandial ; Ayush Chopra ; Mausoom Sarkar ; Piyush Gupta ; Balaji Krishnamurthy ; Vineeth Balasubramanian
COMMENTS:       Accepted at KDD 2020; The first two authors contributed equally
HIGHLIGHT:      In this work, we introduce a new retrospective loss to improve the training of deep neural network models by utilizing the prior experience available in past model states during training.

46, TITLE:      Bringing Light Into the Dark: A Large-scale Evaluation of Knowledge Graph Embedding Models Under a Unified Framework
http://arxiv.org/abs/2006.13365
AUTHORS:        Mehdi Ali ; Max Berrendorf ; Charles Tapley Hoyt ; Laurent Vermue ; Mikhail Galkin ; Sahand Sharifzadeh ; Asja Fischer ; Volker Tresp ; Jens Lehmann
HIGHLIGHT:      We present insights gained as to best practices, best configurations for each model, and where improvements could be made over previously published best configurations.

47, TITLE:      Reducing Overestimation Bias by Increasing Representation Dissimilarity in Ensemble Based Deep Q-Learning
http://arxiv.org/abs/2006.13823
AUTHORS:        Hassam Ullah Sheikh ; Ladislau Bölöni
HIGHLIGHT:      In this paper, we describe a regularization technique to increase the dissimilarity in the representation space in these algorithms.

48, TITLE:      X-ModalNet: A Semi-Supervised Deep Cross-Modal Network for Classification of Remote Sensing Data
http://arxiv.org/abs/2006.13806
AUTHORS:        Danfeng Hong ; Naoto Yokoya ; Gui-Song Xia ; Jocelyn Chanussot ; Xiao Xiang Zhu
HIGHLIGHT:      To this end, we propose a novel cross-modal deep-learning framework, called X-ModalNet, with three well-designed modules: self-adversarial module, interactive learning module, and label propagation module, by learning to transfer more discriminative information from a small-scale hyperspectral image (HSI) into the classification task using a large-scale MSI or SAR data.

49, TITLE:      A Novel Approach for Correcting Multiple Discrete Rigid In-Plane Motions Artefacts in MRI Scans
http://arxiv.org/abs/2006.13804
AUTHORS:        Michael Rotman ; Rafi Brada ; Israel Beniaminy ; Sangtae Ahn ; Christopher J. Hardy ; Lior Wolf
HIGHLIGHT:      In this paper we propose a novel method for removing motion artefacts using a deep neural network with two input branches that discriminates between patient poses using the motion's timing.

50, TITLE:      Interpretable Deep Models for Cardiac Resynchronisation Therapy Response Prediction
http://arxiv.org/abs/2006.13811
AUTHORS:        Esther Puyol-Antón ; Chen Chen ; James R. Clough ; Bram Ruijsink ; Baldeep S. Sidhu ; Justin Gould ; Bradley Porter ; Mark Elliott ; Vishal Mehta ; Daniel Rueckert ; Christopher A. Rinaldi ; Andrew P. King
COMMENTS:       MICCAI 2020 conference
HIGHLIGHT:      In this paper we address both of these issues.

51, TITLE:      Learning Disentangled Representations of Video with Missing Data
http://arxiv.org/abs/2006.13391
AUTHORS:        Armand Comas Massague ; Chi Zhang ; Zlatan Feric ; Octavia Camps ; Rose Yu
HIGHLIGHT:      We present Disentangled Imputed Video autoEncoder (DIVE), a deep generative model that imputes and predicts future video frames in the presence of missing data.

52, TITLE:      Turbocharging Treewidth-Bounded Bayesian Network Structure Learning
http://arxiv.org/abs/2006.13843
AUTHORS:        Vaidyanathan P. R. ; Stefan Szeider
COMMENTS:       12 pages, 1 figure, 2 tables. Source code available at https://www.ac.tuwien.ac.at/files/resources/software/bnslim.zip
HIGHLIGHT:      We present a new approach for learning the structure of a treewidth-bounded Bayesian Network (BN).

53, TITLE:      Comprehensive Information Integration Modeling Framework for Video Titling
http://arxiv.org/abs/2006.13608
AUTHORS:        Shengyu Zhang ; Ziqi Tan ; Jin Yu ; Zhou Zhao ; Kun Kuang ; Tan Jiang ; Jingren Zhou ; Hongxia Yang ; Fei Wu
COMMENTS:       11 pages, 6 figures, to appear in KDD 2020 proceedings
HIGHLIGHT:      To bridge this gap, we integrate comprehensive sources of information, including the content of consumer-generated videos, the narrative comment sentences supplied by consumers, and the product attributes, in an end-to-end modeling framework. We collect a large-scale dataset accordingly from real-world data in Taobao, a world-leading e-commerce platform, and will make the desensitized version publicly available to nourish further development of the research community...

54, TITLE:      Circuit Routing Using Monte Carlo Tree Search and Deep Neural Networks
http://arxiv.org/abs/2006.13607
AUTHORS:        Youbiao He ; Forrest Sheng Bao
HIGHLIGHT:      In this paper, we model the circuit routing as a sequential decision-making problem, and solve it by Monte Carlo tree search (MCTS) with deep neural network (DNN) guided rollout.

55, TITLE:      DeepTracking-Net: 3D Tracking with Unsupervised Learning of Continuous Flow
http://arxiv.org/abs/2006.13848
AUTHORS:        Shuaihang Yuan ; Xiang Li ; Yi Fang
COMMENTS:       16 pages, 6 figures
HIGHLIGHT:      In this paper, we aim at handling the problem of 3D tracking, which provides the tracking of the consecutive frames of 3D shapes.

56, TITLE:      Using Deep Learning and Explainable Artificial Intelligence in Patients' Choices of Hospital Levels
http://arxiv.org/abs/2006.13427
AUTHORS:        Lichin Chen ; Yu Tsao ; Ji-Tian Sheu
COMMENTS:       19 pages, 6 figures
HIGHLIGHT:      Deep learning methods can process highly imbalanced data and achieve high accuracy.

57, TITLE:      A High-Quality Multilingual Dataset for Structured Documentation Translation
http://arxiv.org/abs/2006.13425
AUTHORS:        Kazuma Hashimoto ; Raffaella Buschiazzo ; James Bradbury ; Teresa Marshall ; Richard Socher ; Caiming Xiong
COMMENTS:       Published at WMT2019; the draft has been updated with our dataset's URL: https://github.com/salesforce/localization-xml-mt
HIGHLIGHT:      This paper presents a high-quality multilingual dataset for the documentation domain to advance research on localization of structured text.

58, TITLE:      Recurrent Relational Memory Network for Unsupervised Image Captioning
http://arxiv.org/abs/2006.13611
AUTHORS:        Dan Guo ; Yang Wang ; Peipei Song ; Meng Wang
COMMENTS:       Appearing at IJCAI 2020
HIGHLIGHT:      In this paper, we propose a novel memory-based network rather than GAN, named Recurrent Relational Memory Network ($R^2M$).

59, TITLE:      Explainable robotic systems: Interpreting outcome-focused actions in a reinforcement learning scenario
http://arxiv.org/abs/2006.13615
AUTHORS:        Francisco Cruz ; Richard Dazeley ; Peter Vamplew
COMMENTS:       23 pages, 21 figures
HIGHLIGHT:      In this work, we focus on the decision-making process of a reinforcement learning agent performing a navigation task in a robotic scenario.

60, TITLE:      Movement Tracking by Optical Flow Assisted Inertial Navigation
http://arxiv.org/abs/2006.13856
AUTHORS:        Lassi Meronen ; William J. Wilkinson ; Arno Solin
HIGHLIGHT:      We consider a visually dense approach, where the IMU data is fused with the dense optical flow field estimated from the camera data.

61, TITLE:      GMMLoc: Structure Consistent Visual Localization with Gaussian Mixture Models
http://arxiv.org/abs/2006.13670
AUTHORS:        Huaiyang Huang ; Haoyang Ye ; Yuxiang Sun ; Ming Liu
COMMENTS:       IEEE Robotics and Automation Letters (RA-L); 8 pages, 6 figures
HIGHLIGHT:      In this letter, we aim to address the paradox between accuracy and efficiency in coupling visual factors with structure constraints.

62, TITLE:      GIFnets: Differentiable GIF Encoding Framework
http://arxiv.org/abs/2006.13434
AUTHORS:        Innfarn Yoo ; Xiyang Luo ; Yilin Wang ; Feng Yang ; Peyman Milanfar
HIGHLIGHT:      To reduce artifacts and provide a better and more efficient GIF encoding, we introduce a differentiable GIF encoding pipeline, which includes three novel neural networks: PaletteNet, DitherNet, and BandingNet.

63, TITLE:      AReLU: Attention-based Rectified Linear Unit
http://arxiv.org/abs/2006.13858
AUTHORS:        Dengsheng Chen ; Kai Xu
COMMENTS:       8-page main paper and 6-page appendix
HIGHLIGHT:      We propose a new perspective of learnable activation function through formulating them with element-wise attention mechanism.

64, TITLE:      DINGO: an ontology for projects and grants linked data
http://arxiv.org/abs/2006.13438
AUTHORS:        Diego Chialva ; Alexis-Michel Mugabushaka
COMMENTS:       Accepted for the SKG2020 Workshop co-located with TPDL 2020
HIGHLIGHT:      We present DINGO (Data INtegration for Grants Ontology), an ontology that provides a machine readable extensible framework to model data for semantically-enabled applications relative to projects, funding, actors, and, notably, funding policies in the research landscape.

65, TITLE:      Unifying Optimization Methods for Color Filter Design
http://arxiv.org/abs/2006.13622
AUTHORS:        Graham Finlayson ; Yuteng Zhu
COMMENTS:       11 pages, 3 figures, 1 table
HIGHLIGHT:      In this paper we begin by observing that the function defining the Vora-Value is equivalent to the Luther-condition optimization if we use the orthonormal basis of the XYZ color matching functions, i.e. we linearly transform the XYZ sensitivities to a set of orthonormal basis.

66, TITLE:      Feedback Graph Attention Convolutional Network for Medical Image Enhancement
http://arxiv.org/abs/2006.13863
AUTHORS:        Xiaobin Hu ; Yanyang Yan ; Wenqi Ren ; Hongwei Li ; Yu Zhao ; Amirhossein Bayat ; Bjoern Menze
COMMENTS:       Magnetic resonance imaging and image enhancement and distortions degrading MRI and graph similarity branch and feedback mechanism
HIGHLIGHT:      To well exploit global structural information and texture details, we propose a novel biomedical image enhancement network, named Feedback Graph Attention Convolutional Network (FB-GACN).

67, TITLE:      Hardness of Approximation of (Multi-)LCS over Small Alphabet
http://arxiv.org/abs/2006.13449
AUTHORS:        Amey Bhangale ; Diptarka Chakraborty ; Rajendra Kumar
HIGHLIGHT:      In this paper, we make a significant progress towards proving hardness of approximation over small alphabet by showing a polynomial-time reduction from the well-studied \emph{densest $k$-subgraph} problem with {\em perfect completeness} to approximating Multi-LCS over alphabet of size $poly(n/k)$.

68, TITLE:      OvA-INN: Continual Learning with Invertible Neural Networks
http://arxiv.org/abs/2006.13772
AUTHORS:        G. Hocquet ; O. Bichler ; D. Querlioz
COMMENTS:       to be published in IJCNN 2020
HIGHLIGHT:      In this article, we propose a new method, OvA-INN, which is able to learn one class at a time and without storing any of the previous data.

69, TITLE:      Self-Convolution: A Highly-Efficient Operator for Non-Local Image Restoration
http://arxiv.org/abs/2006.13714
AUTHORS:        Lanqing Guo ; Saiprasad Ravishankar ; Bihan Wen
HIGHLIGHT:      In this work, we propose a novel self-convolution operator to exploit image non-local similarity in a self-supervised way.

70, TITLE:      Disjointness through the Lens of Vapnik-Chervonenkis Dimension: Sparsity and Beyond
http://arxiv.org/abs/2006.13712
AUTHORS:        Anup Bhattacharya ; Sourav Chakraborty ; Arijit Ghosh ; Gopinath Mishra ; Manaswi Paraashar
COMMENTS:       To appear in RANDOM 2020. Pages: 15
HIGHLIGHT:      In this work, we explore how communication complexity measures change with respect to the complexity of the underlying set system.

71, TITLE:      Namira Soccer 2D Simulation Team Description Paper 2020
http://arxiv.org/abs/2006.13534
AUTHORS:        Ehsan Asali ; Farzin Negahbani ; Shahriyar Bamaei ; Zahra Abbasi
HIGHLIGHT:      In this article, we will discuss methods and ideas which are implemented on Namira 2D Soccer Simulation team in the recent year.

72, TITLE:      Benchmark and Best Practices for Biomedical Knowledge Graph Embeddings
http://arxiv.org/abs/2006.13774
AUTHORS:        David Chang ; Ivana Balazevic ; Carl Allen ; Daniel Chawla ; Cynthia Brandt ; Richard Andrew Taylor
COMMENTS:       Accepted to BioNLP 2020 at ACL
HIGHLIGHT:      We train several state-of-the-art knowledge graph embedding models on the SNOMED-CT knowledge graph, provide a benchmark with comparison to existing methods and in-depth discussion on best practices, and make a case for the importance of leveraging the multi-relational nature of knowledge graphs for learning biomedical knowledge representation.

73, TITLE:      Learning Semantically Enhanced Feature for Fine-Grained Image Classification
http://arxiv.org/abs/2006.13457
AUTHORS:        Wei Luo ; Hengmin Zhang ; Jun Li ; Xiu-Shen Wei
COMMENTS:       4 pages, 4 figures
HIGHLIGHT:      We target at providing a computational cheap yet effective approach for fine-grained image classification (FGIC) in this paper.

74, TITLE:      FBK-HUPBA Submission to the EPIC-Kitchens Action Recognition 2020 Challenge
http://arxiv.org/abs/2006.13725
AUTHORS:        Swathikiran Sudhakaran ; Sergio Escalera ; Oswald Lanz
COMMENTS:       Ranked 3rd in the EPIC-Kitchens action recognition challenge @ CVPR 2020
HIGHLIGHT:      In this report we describe the technical details of our submission to the EPIC-Kitchens Action Recognition 2020 Challenge.

75, TITLE:      Imbalanced Gradients: A New Cause of Overestimated Adversarial Robustness
http://arxiv.org/abs/2006.13726
AUTHORS:        Linxi Jiang ; Xingjun Ma ; Zejia Weng ; James Bailey ; Yu-Gang Jiang
HIGHLIGHT:      In this paper, we identify a more subtle situation called \emph{Imbalanced Gradients} that can also cause overestimated adversarial robustness.

76, TITLE:      DeepAbstract: Neural Network Abstraction for Accelerating Verification
http://arxiv.org/abs/2006.13735
AUTHORS:        Pranav Ashok ; Vahid Hashemi ; Jan Křetínský ; Stefanie Mohr
COMMENTS:       Accepted at ATVA 2020
HIGHLIGHT:      We introduce an abstraction framework applicable to fully-connected feed-forward neural networks based on clustering of neurons that behave similarly on some inputs.

77, TITLE:      IA-MOT: Instance-Aware Multi-Object Tracking with Motion Consistency
http://arxiv.org/abs/2006.13458
AUTHORS:        Jiarui Cai ; Yizhou Wang ; Haotian Zhang ; Hung-Min Hsu ; Chengqian Ma ; Jenq-Neng Hwang
COMMENTS:       The 5th Benchmarking Multi-Target Tracking (BMTT) Workshop, CVPR 2020
HIGHLIGHT:      In this work, we propose a novel tracking framework, called "instance-aware MOT" (IA-MOT), that can track multiple objects in either static or moving cameras by jointly considering the instance-level features and object motions.

78, TITLE:      Neural Splines: Fitting 3D Surfaces with Infinitely-Wide Neural Networks
http://arxiv.org/abs/2006.13782
AUTHORS:        Francis Williams ; Matthew Trager ; Joan Bruna ; Denis Zorin
HIGHLIGHT:      We present Neural Splines, a technique for 3D surface reconstruction that is based on random feature kernels arising from infinitely-wide shallow ReLU networks.

79, TITLE:      PhishGAN: Data Augmentation and Identification of Homoglpyh Attacks
http://arxiv.org/abs/2006.13742
AUTHORS:        Joon Sern Lee ; Gui Peng David Yam ; Jin Hao Chan
COMMENTS:       8 pages, 8 figures
HIGHLIGHT:      Here, we show how a conditional Generative Adversarial Network (GAN), PhishGAN, can be used to generate images of hieroglyphs, conditioned on non-homoglpyh input text images.

80, TITLE:      Flexible Image Denoising with Multi-layer Conditional Feature Modulation
http://arxiv.org/abs/2006.13500
AUTHORS:        Jiazhi Du ; Xin Qiao ; Zifei Yan ; Hongzhi Zhang ; Wangmeng Zuo
HIGHLIGHT:      In this paper, we present a novel flexible image enoising network (CFMNet) by equipping an U-Net backbone with multi-layer conditional feature modulation (CFM) modules.

81, TITLE:      Graph Policy Network for Transferable Active Learning on Graphs
http://arxiv.org/abs/2006.13463
AUTHORS:        Shengding Hu ; Zheng Xiong ; Meng Qu ; Xingdi Yuan ; Marc-Alexandre Côté ; Zhiyuan Liu ; Jian Tang
HIGHLIGHT:      In this paper, we study active learning for GNNs, i.e., how to efficiently label the nodes on a graph to reduce the annotation cost of training GNNs.

82, TITLE:      DeepMnemonic: Password Mnemonic Generation via Deep Attentive Encoder-Decoder Model
http://arxiv.org/abs/2006.13462
AUTHORS:        Yao Cheng ; Chang Xu ; Zhen Hai ; Yingjiu Li
COMMENTS:       Published in IEEE Transactions on Dependable and Secure Computing (TDSC)
HIGHLIGHT:      In this paper, we aim to bridge the gap between strong password generation and the usability of strong passwords.

83, TITLE:      ATSO: Asynchronous Teacher-Student Optimizationfor Semi-Supervised Medical Image Segmentation
http://arxiv.org/abs/2006.13461
AUTHORS:        Xinyue Huo ; Lingxi Xie ; Jianzhong He ; Zijie Yang ; Qi Tian
HIGHLIGHT:      This paper focuses on a popular pipeline known as self learning, and points out a weakness namedlazy learningthat refers to the difficulty for a model to learn from the pseudo labels generated by itself.

84, TITLE:      On Analyzing Annotation Consistency in Online Abusive Behavior Datasets
http://arxiv.org/abs/2006.13507
AUTHORS:        Md Rabiul Awal ; Rui Cao ; Roy Ka-Wei Lee ; Sandra Mitrović
HIGHLIGHT:      In this study, we proposed an analytical framework to study the annotation consistency in online hate and abusive content datasets.

85, TITLE:      Insights from the Future for Continual Learning
http://arxiv.org/abs/2006.13748
AUTHORS:        Arthur Douillard ; Eduardo Valle ; Charles Ollion ; Thomas Robert ; Matthieu Cord
HIGHLIGHT:      We introduce Ghost Model, a representation-learning-based model for continual learning using ideas from zero-shot learning.

86, TITLE:      Quantifying Differences in Reward Functions
http://arxiv.org/abs/2006.13900
AUTHORS:        Adam Gleave ; Michael Dennis ; Shane Legg ; Stuart Russell ; Jan Leike
COMMENTS:       8 pages main paper, 29 pages total
HIGHLIGHT:      To address these problems, we introduce the Equivalent-Policy Invariant Comparison (EPIC) distance to quantify the difference between two reward functions directly, without training a policy.

87, TITLE:      Feature-dependent Cross-Connections in Multi-Path Neural Networks
http://arxiv.org/abs/2006.13904
AUTHORS:        Dumindu Tissera ; Kasun Vithanage ; Rukshan Wijesinghe ; Kumara Kahatapitiya ; Subha Fernando ; Ranga Rodrigo
COMMENTS:       Accepted to ICPR 2020
HIGHLIGHT:      To do this, we propose inserting feature-dependent cross-connections between parallel sets of feature maps in successive layers.

88, TITLE:      Disentangle Perceptual Learning through Online Contrastive Learning
http://arxiv.org/abs/2006.13511
AUTHORS:        Kangfu Mei ; Yao Lu ; Qiaosi Yi ; Haoyu Wu ; Juncheng Li ; Rui Huang
COMMENTS:       12 pages, 8 figures
HIGHLIGHT:      In this paper, we argue that, among the features representation from the pre-trained classification network, only limited dimensions are related to human visual perception, while others are irrelevant, although both will affect the final image transformation results.

89, TITLE:      Dynamic Functional Connectivity and Graph Convolution Network for Alzheimer's Disease Classification
http://arxiv.org/abs/2006.13510
AUTHORS:        Xingwei An ; Yutao Zhou ; Yang Di ; Dong Ming
HIGHLIGHT:      In this paper, we introduce a novel method based on dynamic functional connectivity (dFC) that can effectively capture changes in the brain.

90, TITLE:      Black-box Adaptation of ASR for Accented Speech
http://arxiv.org/abs/2006.13519
AUTHORS:        Kartik Khandelwal ; Preethi Jyothi ; Abhijeet Awasthi ; Sunita Sarawagi
COMMENTS:       A slightly different version submitted to INTERSPEECH 2020 (currently under review)
HIGHLIGHT:      We introduce the problem of adapting a black-box, cloud-based ASR system to speech from a target accent.

91, TITLE:      3DMotion-Net: Learning Continuous Flow Function for 3D Motion Prediction
http://arxiv.org/abs/2006.13906
AUTHORS:        Shuaihang Yuan ; Xiang Li ; Anthony Tzes ; Yi Fang
COMMENTS:       8 pages, 7 figures
HIGHLIGHT:      In this paper, we deal with the problem to predict the future 3D motions of 3D object scans from previous two consecutive frames.

92, TITLE:      3D Pose Detection in Videos: Focusing on Occlusion
http://arxiv.org/abs/2006.13517
AUTHORS:        Justin Wang ; Edward Xu ; Kangrui Xue ; Lukasz Kidzinski
HIGHLIGHT:      In this work, we build upon existing methods for occlusion-aware 3D pose detection in videos.

93, TITLE:      The NetHack Learning Environment
http://arxiv.org/abs/2006.13760
AUTHORS:        Heinrich Küttler ; Nantas Nardelli ; Alexander H. Miller ; Roberta Raileanu ; Marco Selvatici ; Edward Grefenstette ; Tim Rocktäschel
COMMENTS:       27 pages
HIGHLIGHT:      Here, we present the NetHack Learning Environment (NLE), a scalable, procedurally generated, stochastic, rich, and challenging environment for RL research based on the popular single-player terminal-based roguelike game, NetHack.

94, TITLE:      Off-Dynamics Reinforcement Learning: Training for Transfer with Domain Classifiers
http://arxiv.org/abs/2006.13916
AUTHORS:        Benjamin Eysenbach ; Swapnil Asawa ; Shreyas Chaudhari ; Ruslan Salakhutdinov ; Sergey Levine
HIGHLIGHT:      We propose a simple, practical, and intuitive approach for domain adaptation in reinforcement learning.

95, TITLE:      Generative causal explanations of black-box classifiers
http://arxiv.org/abs/2006.13913
AUTHORS:        Matthew O'Shaughnessy ; Gregory Canal ; Marissa Connor ; Mark Davenport ; Christopher Rozell
COMMENTS:       12+16 pages, 5+11 figures
HIGHLIGHT:      Using carefully controlled test cases, we provide intuition that illuminates the function of our causal objective.

96, TITLE:      Improving task-specific representation via 1M unlabelled images without any extra knowledge
http://arxiv.org/abs/2006.13919
AUTHORS:        Aayush Bansal
COMMENTS:       Technical Report
HIGHLIGHT:      We propose an exceedingly simple method of conditioning an existing representation on a diverse data distribution and observe that a model trained on diverse examples acts as a better initialization.

97, TITLE:      Competitive Balance in Team Sports Games
http://arxiv.org/abs/2006.13763
AUTHORS:        Sofia M Nikolakaki ; Ogheneovo Dibie ; Ahmad Beirami ; Nicholas Peterson ; Navid Aghdaie ; Kazi Zaman
COMMENTS:       2020 IEEE Conference in Games (COG 2020), 8 pages
HIGHLIGHT:      In this paper, we show that using final score difference provides yet a better prediction metric for competitive balance.

98, TITLE:      Towards Adversarial Planning for Indoor Scenes with Rotation
http://arxiv.org/abs/2006.13527
AUTHORS:        Xinhan Di ; Pengqian Yu ; Hong Zhu ; Lei Cai ; Qiuyan Sheng ; Changyu Sun
COMMENTS:       submit to conference
HIGHLIGHT:      In this paper, we propose an adversarial model for producing furniture layout for interior scene synthesis when the interior room is rotated. We provide an interior layout dataset that contains $14400$ designs from the professional designers with rotation.



==========Updates to Previous Papers==========
1, TITLE:       Unsupervised Discovery, Control, and Disentanglement of Semantic Attributes with Applications to Anomaly Detection
http://arxiv.org/abs/2002.11169
AUTHORS:        William Paul ; I-Jeng Wang ; Fady Alajaji ; Philippe Burlina
COMMENTS:       18 pages, 5 figures, preprint
HIGHLIGHT:      For (a) we propose a network architecture that exploits the combination of multiscale generative models with mutual information (MI) maximization.

2, TITLE:       hxtorch: PyTorch for BrainScaleS-2 -- Perceptrons on Analog Neuromorphic Hardware
http://arxiv.org/abs/2006.13138
AUTHORS:        Philipp Spilger ; Eric Müller ; Arne Emmel ; Aron Leibfried ; Christian Mauch ; Christian Pehle ; Johannes Weis ; Oliver Breitwieser ; Sebastian Billaudelle ; Sebastian Schmitt ; Timo C. Wunderlich ; Yannik Stradmann ; Johannes Schemmel
HIGHLIGHT:      As an application of the introduced framework, we present a model that classifies activities of daily living with smartphone sensor data.

3, TITLE:       Vatex Video Captioning Challenge 2020: Multi-View Features and Hybrid Reward Strategies for Video Captioning
http://arxiv.org/abs/1910.11102
AUTHORS:        Xinxin Zhu ; Longteng Guo ; Peng Yao ; Shichen Lu ; Wei Liu ; Jing Liu
COMMENTS:       4 pages,2 figure
HIGHLIGHT:      This report describes our solution for the VATEX Captioning Challenge 2020, which requires generating descriptions for the videos in both English and Chinese languages.

4, TITLE:       Implicit Generation and Modeling in Energy-Based Models
http://arxiv.org/abs/1903.08689
AUTHORS:        Yilun Du ; Igor Mordatch
HIGHLIGHT:      We present techniques to scale MCMC based EBM training on continuous neural networks, and we show its success on the high-dimensional data domains of ImageNet32x32, ImageNet128x128, CIFAR-10, and robotic hand trajectories, achieving better samples than other likelihood models and nearing the performance of contemporary GAN approaches, while covering all modes of the data.

5, TITLE:       TextGAIL: Generative Adversarial Imitation Learning for Text Generation
http://arxiv.org/abs/2004.13796
AUTHORS:        Qingyang Wu ; Lei Li ; Zhou Yu
HIGHLIGHT:      To address this problem, we propose a generative adversarial imitation learning framework for text generation that uses large pre-trained language models to provide more reliable reward guidance.

6, TITLE:       Learning medical triage from clinicians using Deep Q-Learning
http://arxiv.org/abs/2003.12828
AUTHORS:        Albert Buchard ; Baptiste Bouvier ; Giulia Prando ; Rory Beard ; Michail Livieratos ; Dan Busbridge ; Daniel Thompson ; Jonathan Richens ; Yuanzhao Zhang ; Adam Baker ; Yura Perov ; Kostis Gourgoulias ; Saurabh Johri
COMMENTS:       17 pages, 4 figures, 3 tables, preprint, in press
HIGHLIGHT:      In this work, we present a Deep Reinforcement Learning approach (a variant of DeepQ-Learning) to triage patients using curated clinical vignettes.

7, TITLE:       CBR-Net: Cascade Boundary Refinement Network for Action Detection: Submission to ActivityNet Challenge 2020 (Task 1)
http://arxiv.org/abs/2006.07526
AUTHORS:        Xiang Wang ; Baiteng Ma ; Zhiwu Qing ; Yongpeng Sang ; Changxin Gao ; Shiwei Zhang ; Nong Sang
COMMENTS:       ActivityNet Challenge 2020 Temporal Action Localization (Task 1) Champion Solution (Rank 1)
HIGHLIGHT:      In this report, we present our solution for the task of temporal action localization (detection) (task 1) in ActivityNet Challenge 2020.

8, TITLE:       Modular Termination for Second-Order Computation Rules and Application to Algebraic Effect Handlers
http://arxiv.org/abs/1912.03434
AUTHORS:        Makoto Hamana
COMMENTS:       26 pages
HIGHLIGHT:      We present a new modular proof method of termination for second-order computation, and report its implementation SOL.

9, TITLE:       Variational Model-based Policy Optimization
http://arxiv.org/abs/2006.05443
AUTHORS:        Yinlam Chow ; Brandon Cui ; MoonKyung Ryu ; Mohammad Ghavamzadeh
HIGHLIGHT:      In this paper, we leverage the connection between RL and probabilistic inference, and formulate such an objective function as a variational lower-bound of a log-likelihood.

10, TITLE:      LIBRE: The Multiple 3D LiDAR Dataset
http://arxiv.org/abs/2003.06129
AUTHORS:        Alexander Carballo ; Jacob Lambert ; Abraham Monrroy-Cano ; David Robert Wong ; Patiphon Narksri ; Yuki Kitsukawa ; Eijiro Takeuchi ; Shinpei Kato ; Kazuya Takeda
COMMENTS:       Accepted for oral presentation at IEEE Intelligent Vehicles Symposium (IV2020), https://2020.ieee-iv.org/ LIBRE dataset available at https://sites.google.com/g.sp.m.is.nagoya-u.ac.jp/libre-dataset/ Reference video available at https://youtu.be/rWyecoCtKcQ
HIGHLIGHT:      In this work, we present LIBRE: LiDAR Benchmarking and Reference, a first-of-its-kind dataset featuring 10 different LiDAR sensors, covering a range of manufacturers, models, and laser configurations.

11, TITLE:      Optimizing Data Usage via Differentiable Rewards
http://arxiv.org/abs/1911.10088
AUTHORS:        Xinyi Wang ; Hieu Pham ; Paul Michel ; Antonios Anastasopoulos ; Jaime Carbonell ; Graham Neubig
COMMENTS:       Accepted at ICML 2020
HIGHLIGHT:      To efficiently optimize data usage, we propose a reinforcement learning approach called Differentiable Data Selection (DDS).

12, TITLE:      A Multi-Turn Emotionally Engaging Dialog Model
http://arxiv.org/abs/1908.07816
AUTHORS:        Yubo Xie ; Ekaterina Svikhnushina ; Pearl Pu
COMMENTS:       Accepted to IUI 2020 user2agent workshop
HIGHLIGHT:      In this paper, we propose a multi-turn dialog system aimed at learning and generating emotional responses that so far only humans know how to do.

13, TITLE:      Generating Annotated High-Fidelity Images Containing Multiple Coherent Objects
http://arxiv.org/abs/2006.12150
AUTHORS:        Bryan G. Cardenas ; Devanshu Arya ; Deepak K. Gupta
COMMENTS:       21 pages, 5 tables, 21 figures
HIGHLIGHT:      In this work, we propose a multi-object generation framework that can synthesize images with multiple objects without explicitly requiring their contextual information during the generation process.

14, TITLE:      Message Passing Query Embedding
http://arxiv.org/abs/2002.02406
AUTHORS:        Daniel Daza ; Michael Cochez
COMMENTS:       Presented at ICML 2020 - GRL+ Workshop
HIGHLIGHT:      We propose a more general architecture that employs a graph neural network to encode a graph representation of the query, where nodes correspond to entities and variables.

15, TITLE:      Coreference Resolution as Query-based Span Prediction
http://arxiv.org/abs/1911.01746
AUTHORS:        Wei Wu ; Fei Wang ; Arianna Yuan ; Fei Wu ; Jiwei Li
HIGHLIGHT:      In this paper, we present an accurate and extensible approach for the coreference resolution task.

16, TITLE:      FairALM: Augmented Lagrangian Method for Training Fair Models with Little Regret
http://arxiv.org/abs/2004.01355
AUTHORS:        Vishnu Suresh Lokhande ; Aditya Kumar Akash ; Sathya N. Ravi ; Vikas Singh
HIGHLIGHT:      It is now accepted that because of biases in the datasets we present to the models, a fairness-oblivious training will lead to unfair models.

17, TITLE:      Pruning untrained neural networks: Principles and Analysis
http://arxiv.org/abs/2002.08797
AUTHORS:        Soufiane Hayou ; Jean-Francois Ton ; Arnaud Doucet ; Yee Whye Teh
COMMENTS:       31 pages, 12 figures
HIGHLIGHT:      In this paper we provide a comprehensive theoretical analysis of pruning at initialization and training of sparse architectures.

18, TITLE:      A Communication Efficient Collaborative Learning Framework for Distributed Features
http://arxiv.org/abs/1912.11187
AUTHORS:        Yang Liu ; Yan Kang ; Xinwei Zhang ; Liping Li ; Yong Cheng ; Tianjian Chen ; Mingyi Hong ; Qiang Yang
COMMENTS:       This paper is published at the 2nd International Workshop on Federated Learning for Data Privacy and Confidentiality, in Conjunction with NeurIPS 2019 (FL-NeurIPS 19): https://nips.cc/Conferences/2019/ScheduleMultitrack?event=13202
HIGHLIGHT:      We introduce a collaborative learning framework allowing multiple parties having different sets of attributes about the same user to jointly build models without exposing their raw data or model parameters.

19, TITLE:      Sentence Meta-Embeddings for Unsupervised Semantic Textual Similarity
http://arxiv.org/abs/1911.03700
AUTHORS:        Nina Poerner ; Ulli Waltinger ; Hinrich Schütze
HIGHLIGHT:      We apply, extend and evaluate different meta-embedding methods from the word embedding literature at the sentence level, including dimensionality reduction (Yin and Sch\"utze, 2016), generalized Canonical Correlation Analysis (Rastogi et al., 2015) and cross-view auto-encoders (Bollegala and Bao, 2018).

20, TITLE:      Detection in Crowded Scenes: One Proposal, Multiple Predictions
http://arxiv.org/abs/2003.09163
AUTHORS:        Xuangeng Chu ; Anlin Zheng ; Xiangyu Zhang ; Jian Sun
COMMENTS:       CVPR 2020 Oral
HIGHLIGHT:      We propose a simple yet effective proposal-based object detector, aiming at detecting highly-overlapped instances in crowded scenes.

21, TITLE:      DCNNs: A Transfer Learning comparison of Full Weapon Family threat detection for Dual-Energy X-Ray Baggage Imagery
http://arxiv.org/abs/2006.13065
AUTHORS:        A. Williamson ; P. Dickinson ; T. Lambrou ; J. C. Murray
COMMENTS:       Submitted to BMVC 2019 Workshop on "Object Detection and Recognition for Security Screening"
HIGHLIGHT:      In this work we propose the first pipeline to effectively process Dual-Energy X-Ray scanner output, and perform classification capable of distinguishing between firearm families (Assault Rifle, Revolver, Self-Loading Pistol,Shotgun, and Sub-Machine Gun) from this output.

22, TITLE:      Adaptive Factorization Network: Learning Adaptive-Order Feature Interactions
http://arxiv.org/abs/1909.03276
AUTHORS:        Weiyu Cheng ; Yanyan Shen ; Linpeng Huang
COMMENTS:       Accepted by AAAI'20
HIGHLIGHT:      In this work, we propose the Adaptive Factorization Network (AFN), a new model that learns arbitrary-order cross features adaptively from data.

23, TITLE:      Extracting COVID-19 Events from Twitter
http://arxiv.org/abs/2006.02567
AUTHORS:        Shi Zong ; Ashutosh Baheti ; Wei Xu ; Alan Ritter
HIGHLIGHT:      We present a corpus of 7,500 tweets annotated with COVID-19 events, including positive test results, denied access to testing, and more.

24, TITLE:      Option Encoder: A Framework for Discovering a Policy Basis in Reinforcement Learning
http://arxiv.org/abs/1909.04134
AUTHORS:        Arjun Manoharan ; Rahul Ramesh ; Balaraman Ravindran
COMMENTS:       Accepted at ECML-PKDD 2020
HIGHLIGHT:      In this work, we propose Option Encoder, an auto-encoder based framework with intelligently constrained weights, that helps discover a collection of basis policies.

25, TITLE:      Learning pronunciation from a foreign language in speech synthesis networks
http://arxiv.org/abs/1811.09364
AUTHORS:        Younggun Lee ; Suwon Shon ; Taesu Kim
HIGHLIGHT:      In this study, we are interested in analyzing and taking advantage of multilingual speech synthesis network.

26, TITLE:      Feel The Music: Automatically Generating A Dance For An Input Song
http://arxiv.org/abs/2006.11905
AUTHORS:        Purva Tendulkar ; Abhishek Das ; Aniruddha Kembhavi ; Devi Parikh
COMMENTS:       4 pages
HIGHLIGHT:      We present a general computational approach that enables a machine to generate a dance for any input music.

27, TITLE:      Investigating Distributional Robustness: Semantic Perturbations Using Generative Models
http://arxiv.org/abs/2001.11055
AUTHORS:        Isaac Dunn ; Laura Hanu ; Hadrien Pouget ; Daniel Kroening ; Tom Melham
COMMENTS:       Updated to include new results
HIGHLIGHT:      In this paper, we introduce a new method for perturbing the semantic features of images (e.g. shape, location, texture, and colour) for the purpose of evaluating classifiers' robustness to these changes.

28, TITLE:      Locally Masked Convolution for Autoregressive Models
http://arxiv.org/abs/2006.12486
AUTHORS:        Ajay Jain ; Pieter Abbeel ; Deepak Pathak
COMMENTS:       Published at Conference on Uncertainty in AI (UAI) 2020
HIGHLIGHT:      To generate data in arbitrary orders, we introduce LMConv: a simple modification to the standard 2D convolution that allows arbitrary masks to be applied to the weights at each location in the image.

29, TITLE:      Keep Your AI-es on the Road: Tackling Distracted Driver Detection with Convolutional Neural Networks and Targeted Data Augmentation
http://arxiv.org/abs/2006.10955
AUTHORS:        Nikka Mofid ; Jasmine Bayrooti ; Shreya Ravi
COMMENTS:       10 pages, 11 figures
HIGHLIGHT:      In our study, we tackle the problem of distracted driving by aiming to build a robust multi-class classifier to detect and identify different forms of driver inattention using the State Farm Distracted Driving Dataset.

30, TITLE:      Rethinking the Trigger of Backdoor Attack
http://arxiv.org/abs/2004.04692
AUTHORS:        Yiming Li ; Tongqing Zhai ; Baoyuan Wu ; Yong Jiang ; Zhifeng Li ; Shutao Xia
COMMENTS:       13 pages
HIGHLIGHT:      In this paper, we start with the study of the property of the backdoor trigger.

31, TITLE:      A Qualitative Evaluation of Language Models on Automatic Question-Answering for COVID-19
http://arxiv.org/abs/2006.10964
AUTHORS:        David Oniani ; Yanshan Wang
HIGHLIGHT:      In this paper, we propose to apply a language model for automatically answering questions related to COVID-19 and qualitatively evaluate the generated responses.

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

33, TITLE:      GhostImage: Remote Perception Attacks against Camera-based Image Classification Systems
http://arxiv.org/abs/2001.07792
AUTHORS:        Yanmao Man ; Ming Li ; Ryan Gerdes
COMMENTS:       Accepted by USENIX RAID 2020. Source code is available at https://github.com/Harry1993/GhostImage
HIGHLIGHT:      In this work we demonstrate how the perception domain can be remotely and unobtrusively exploited to enable an attacker to create spurious objects or alter an existing object.

34, TITLE:      Event Representation Learning Enhanced with External Commonsense Knowledge
http://arxiv.org/abs/1909.05190
AUTHORS:        Xiao Ding ; Kuo Liao ; Ting Liu ; Zhongyang Li ; Junwen Duan
HIGHLIGHT:      Prior work has proposed effective methods to learn event representations that can capture syntactic and semantic information over text corpus, demonstrating their effectiveness for downstream tasks such as script event prediction.

35, TITLE:      On the Relationship Between Active Inference and Control as Inference
http://arxiv.org/abs/2006.12964
AUTHORS:        Beren Millidge ; Alexander Tschantz ; Anil K Seth ; Christopher L Buckley
COMMENTS:       initial upload
HIGHLIGHT:      In the context of this comparison, we highlight several ways in which these frameworks can inform one another.

36, TITLE:      Proving Data-Poisoning Robustness in Decision Trees
http://arxiv.org/abs/1912.00981
AUTHORS:        Samuel Drews ; Aws Albarghouthi ; Loris D'Antoni
COMMENTS:       Changes: revisions to main text for clarity of presentation, and corrections to proofs in the appendices
HIGHLIGHT:      We present a sound verification technique based on abstract interpretation and implement it in a tool called Antidote.

37, TITLE:      ViPR: Visual-Odometry-aided Pose Regression for 6DoF Camera Localization
http://arxiv.org/abs/1912.08263
AUTHORS:        Felix Ott ; Tobias Feigl ; Christoffer Löffler ; Christopher Mutschler
COMMENTS:       Conf. on Computer Vision and Pattern Recognition (CVPR): Joint Workshop on Long-Term Visual Localization, Visual Odometry and Geometric and Learning-based SLAM 2020
HIGHLIGHT:      We propose ViPR, a novel modular architecture for long-term 6DoF VO that leverages temporal information and synergies between absolute pose estimates (from PoseNet-like modules) and relative pose estimates (from FlowNet-based modules) by combining both through recurrent layers.

38, TITLE:      Taming Pretrained Transformers for Extreme Multi-label Text Classification
http://arxiv.org/abs/1905.02331
AUTHORS:        Wei-Cheng Chang ; Hsiang-Fu Yu ; Kai Zhong ; Yiming Yang ; Inderjit Dhillon
COMMENTS:       KDD 2020 Applied Data Track
HIGHLIGHT:      In this paper, we propose X-Transformer, the first scalable approach to fine-tuning deep transformer models for the XMC problem.

39, TITLE:      Revisiting Regex Generation for Modeling Industrial Applications by Incorporating Byte Pair Encoder
http://arxiv.org/abs/2005.02558
AUTHORS:        Desheng Wang ; Jiawei Liu ; Xiang Qi ; Baolin Sun ; Peng Zhang
HIGHLIGHT:      This work focuses on automatically generating regular expressions and proposes a novel genetic algorithm to deal with this problem.

40, TITLE:      On Bayesian Search for the Feasible Space Under Computationally Expensive Constraints
http://arxiv.org/abs/2004.11055
AUTHORS:        Alma Rahat ; Michael Wood
COMMENTS:       Accepted at The Sixth International Conference on Machine Learning, Optimization, and Data Science. Main content 12 pages, a total of 19 pages with supplementary. 3 Figures and 2 tables. Python code for Bayesian search is available at: http://bitbucket.org/arahat/lod-2020
HIGHLIGHT:      In this paper, we propose a novel acquisition function that combines the probability that a solution lies at the boundary between feasible and infeasible spaces (representing exploitation) and the entropy in predictions (representing exploration).

41, TITLE:      The Absent-Minded Passengers Problem: A Motivating Challenge Solved by Computer Algebra
http://arxiv.org/abs/2003.01921
AUTHORS:        Carsten Schneider
COMMENTS:       Removed various typos and inserted an extra link for a Mathematica notebook to repeat the (not-so-costly) calculations
HIGHLIGHT:      In this note we report on this enterprise.

42, TITLE:      Text Complexity Classification Based on Linguistic Information: Application to Intelligent Tutoring of ESL
http://arxiv.org/abs/2001.01863
AUTHORS:        M. Zakaria Kurdi
COMMENTS:       This is an unpublished pre-print, the JDMDH journal requires submission to arxiv.org before the submission to the journal (see the link: https://jdmdh.episciences.org/page/submissions#)
HIGHLIGHT:      The goal of this work is to build a classifier that can identify text complexity within the context of teaching reading to English as a Second Language (ESL) learners.

43, TITLE:      Printing and Scanning Attack for Image Counter Forensics
http://arxiv.org/abs/2005.02160
AUTHORS:        Hailey James ; Otkrist Gupta ; Dan Raviv
COMMENTS:       10 pages, 5 figures, 7 tables
HIGHLIGHT:      In this paper we explore another method of highly plausible attack: printing and scanning. To facilitate this exploration, we create a dataset of over 6,000 printed and scanned image blocks.

44, TITLE:      IQA: Interactive Query Construction in Semantic Question Answering Systems
http://arxiv.org/abs/2006.11534
AUTHORS:        Hamid Zafar ; Mohnish Dubey ; Jens Lehmann ; Elena Demidova
HIGHLIGHT:      In this article, we aim to empower users in guiding QA systems towards the intended semantic queries by means of interaction.

45, TITLE:      Unsupervised Discovery of Interpretable Directions in the GAN Latent Space
http://arxiv.org/abs/2002.03754
AUTHORS:        Andrey Voynov ; Artem Babenko
HIGHLIGHT:      In this paper, we introduce an unsupervised method to identify interpretable directions in the latent space of a pretrained GAN model.

46, TITLE:      A Baseline Approach for AutoImplant: the MICCAI 2020 Cranial Implant Design Challenge
http://arxiv.org/abs/2006.12449
AUTHORS:        Jianning Li ; Antonio Pepe ; Christina Gsaxner ; Gord von Campe ; Jan Egger
COMMENTS:       12 pages
HIGHLIGHT:      In this study, we present a baseline approach for AutoImplant (https://autoimplant.grand-challenge.org/) - the cranial implant design challenge, which, as suggested by the organizers, can be formulated as a volumetric shape learning task.

47, TITLE:      Meta-Learned Confidence for Few-shot Learning
http://arxiv.org/abs/2002.12017
AUTHORS:        Seong Min Kye ; Hae Beom Lee ; Hoirin Kim ; Sung Ju Hwang
HIGHLIGHT:      To tackle this issue, we propose to meta-learn the confidence for each query sample, to assign optimal weights to unlabeled queries such that they improve the model's transductive inference performance on unseen tasks.

48, TITLE:      RP2K: A Large-Scale Retail Product Dataset for Fine-Grained Image Classification
http://arxiv.org/abs/2006.12634
AUTHORS:        Jingtian Peng ; Chang Xiao ; Xun Wei ; Yifan Li
HIGHLIGHT:      We introduce RP2K, a new large-scale retail product dataset for fine-grained image classification.

49, TITLE:      Exploring Software Naturalness through Neural Language Models
http://arxiv.org/abs/2006.12641
AUTHORS:        Luca Buratti ; Saurabh Pujar ; Mihaela Bornea ; Scott McCarley ; Yunhui Zheng ; Gaetano Rossiello ; Alessandro Morari ; Jim Laredo ; Veronika Thost ; Yufan Zhuang ; Giacomo Domeniconi
HIGHLIGHT:      To achieve this, we introduce a sequence labeling task that directly probes the language models understanding of AST.

50, TITLE:      Scribble2Label: Scribble-Supervised Cell Segmentation via Self-Generating Pseudo-Labels with Consistency
http://arxiv.org/abs/2006.12890
AUTHORS:        Hyeonsoo Lee ; Won-Ki Jeong
COMMENTS:       MICCAI 2020 accepted
HIGHLIGHT:      In this paper, we introduce Scribble2Label, a novel weakly-supervised cell segmentation framework that exploits only a handful of scribble annotations without full segmentation labels.

51, TITLE:      Improved Deep Point Cloud Geometry Compression
http://arxiv.org/abs/2006.09043
AUTHORS:        Maurice Quach ; Giuseppe Valenzise ; Frederic Dufaux
COMMENTS:       Code is available at https://github.com/mauriceqch/pcc_geo_cnn_v2
HIGHLIGHT:      In this paper, we propose a set of contributions to improve deep point cloud compression, i.e.: using a scale hyperprior model for entropy coding; employing deeper transforms; a different balancing weight in the focal loss; optimal thresholding for decoding; and sequential model training.

52, TITLE:      Learning Physical Graph Representations from Visual Scenes
http://arxiv.org/abs/2006.12373
AUTHORS:        Daniel M. Bear ; Chaofei Fan ; Damian Mrowca ; Yunzhu Li ; Seth Alter ; Aran Nayebi ; Jeremy Schwartz ; Li Fei-Fei ; Jiajun Wu ; Joshua B. Tenenbaum ; Daniel L. K. Yamins
COMMENTS:       23 pages; corrected affiliations and acknowledgments
HIGHLIGHT:      To overcome these limitations, we introduce the idea of Physical Scene Graphs (PSGs), which represent scenes as hierarchical graphs, with nodes in the hierarchy corresponding intuitively to object parts at different scales, and edges to physical connections between parts.

53, TITLE:      Dissimilarity Mixture Autoencoder for Deep Clustering
http://arxiv.org/abs/2006.08177
AUTHORS:        Juan S. Lara ; Fabio A. González
COMMENTS:       8 pages (5 additional pages for broader impact, references and supplementary material)
HIGHLIGHT:      In this paper, we introduce the Dissimilarity Mixture Autoencoder (DMAE), a novel neural network model that uses a dissimilarity function to generalize a family of density estimation and clustering methods.

54, TITLE:      Chessboard and chess piece recognition with the support of neural networks
http://arxiv.org/abs/1708.03898
AUTHORS:        Maciej A. Czyzewski ; Artur Laskowski ; Szymon Wasik
COMMENTS:       11 pages, 14 figures; for implementation, see https://github.com/maciejczyzewski/neural-chessboard; Submitted to FCDS, In Review
HIGHLIGHT:      To solve this problem, we propose a novel algorithm for digitizing chessboard configurations.

55, TITLE:      End-to-End Face Parsing via Interlinked Convolutional Neural Networks
http://arxiv.org/abs/2002.04831
AUTHORS:        Zi Yin ; Valentin Yiu ; Xiaolin Hu ; Liang Tang
HIGHLIGHT:      To solve this problem, we introduce a simple, end-to-end face parsing framework: STN-aided iCNN(STN-iCNN), which extends the iCNN by adding a Spatial Transformer Network (STN) between the two isolated stages.

56, TITLE:      FMix: Enhancing Mixed Sample Data Augmentation
http://arxiv.org/abs/2002.12047
AUTHORS:        Ethan Harris ; Antonia Marcu ; Matthew Painter ; Mahesan Niranjan ; Adam Prügel-Bennett ; Jonathon Hare
COMMENTS:       16 pages, code available at https://github.com/ecs-vlc/FMix
HIGHLIGHT:      From insight on the efficacy of CutMix in particular, we propose FMix, an MSDA that uses binary masks obtained by applying a threshold to low frequency images sampled from Fourier space.