2020.04.08.txt

==========New Papers==========
1, TITLE:       MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices
http://arxiv.org/abs/2004.02984
AUTHORS:        Zhiqing Sun ; Hongkun Yu ; Xiaodan Song ; Renjie Liu ; Yiming Yang ; Denny Zhou
COMMENTS:       Accepted to ACL 2020
HIGHLIGHT:      In this paper, we propose MobileBERT for compressing and accelerating the popular BERT model.

2, TITLE:       Zero-Shot Learning of Text Adventure Games with Sentence-Level Semantics
http://arxiv.org/abs/2004.02986
AUTHORS:        Xusen Yin ; Jonathan May
HIGHLIGHT:      We introduce a new model amenable to deep Q-learning that incorporates a Siamese neural network architecture and a novel refactoring of the Q-value function in order to better represent system state given its approximation over a language channel.

3, TITLE:       Probabilistic Diagnostic Tests for Degradation Problems in Supervised Learning
http://arxiv.org/abs/2004.02988
AUTHORS:        Gustavo A. Valencia-Zapata ; Okan Ersoy ; Carolina Gonzalez-Canas ; Michael G. Zentner ; Gerhard Klimeck
HIGHLIGHT:      Probabilistic Diagnostic Tests for Degradation Problems in Supervised Learning

4, TITLE:       Evaluating the Evaluation of Diversity in Natural Language Generation
http://arxiv.org/abs/2004.02990
AUTHORS:        Guy Tevet ; Jonathan Berant
HIGHLIGHT:      In this work, we propose a framework for evaluating diversity metrics.

5, TITLE:       Predicting Strategic Behavior from Free Text
http://arxiv.org/abs/2004.02973
AUTHORS:        Omer Ben-Porat ; Sharon Hirsch ; Lital Kuchy ; Guy Elad ; Roi Reichart ; Moshe Tennenholtz
COMMENTS:       Accepted to Journal of Artificial Intelligence Research (JAIR), 2020
HIGHLIGHT:      This paper aims to connect these two strands of research, which we consider highly timely and important due to the vast online textual communication on the web.

6, TITLE:       LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihood
http://arxiv.org/abs/2004.02980
AUTHORS:        Abhinav Kumar ; Tim K. Marks ; Wenxuan Mou ; Ye Wang ; Michael Jones ; Anoop Cherian ; Toshiaki Koike-Akino ; Xiaoming Liu ; Chen Feng
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      In this paper, we present a novel framework for jointly predicting landmark locations, associated uncertainties of these predicted locations, and landmark visibilities.

7, TITLE:       Integrating Owicki-Gries for C11-Style Memory Models into Isabelle/HOL
http://arxiv.org/abs/2004.02983
AUTHORS:        Sadegh Dalvandi ; Brijesh Dongol ; Simon Doherty
HIGHLIGHT:      The technique introduces a set of high-level assertions over C11 states together with a set of basic Hoare-style axioms over atomic weak memory statements (e.g., reads/writes), but retains all other standard proof obligations for compound statements.

8, TITLE:       Multi-Step Inference for Reasoning Over Paragraphs
http://arxiv.org/abs/2004.02995
AUTHORS:        Jiangming Liu ; Matt Gardner
HIGHLIGHT:      We present a middle ground between these two extremes: a compositional model reminiscent of neural module networks that can perform chained logical reasoning.

9, TITLE:       Emergent Language Generalization and Acquisition Speed are not tied to Compositionality
http://arxiv.org/abs/2004.03420
AUTHORS:        Eugene Kharitonov ; Marco Baroni
HIGHLIGHT:      We argue that these beneficial properties are only loosely connected to compositionality.

10, TITLE:      Operationalizing the legal concept of 'Incitement to Hatred' as an NLP task
http://arxiv.org/abs/2004.03422
AUTHORS:        Frederike Zufall ; Huangpan Zhang ; Katharina Kloppenborg ; Torsten Zesch
HIGHLIGHT:      In this paper, we pursue these questions based on the criminal offense of 'incitement to hatred' in {\S} 130 of the German Criminal Code along with the underlying EU Council Framework Decision.

11, TITLE:      Automated Smartphone based System for Diagnosis of Diabetic Retinopathy
http://arxiv.org/abs/2004.03408
AUTHORS:        Misgina Tsighe Hagos ; Shri Kant ; Surayya Ado Bala
COMMENTS:       12 pages, 4 figures, 4 tables, 1 appendix. Copyright \copyright 2019, IEEE. Published in: 2019 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS)
HIGHLIGHT:      In this paper, inception based convolutional neural network and binary decision tree-based ensemble of classifiers have been proposed and implemented to detect and classify diabetic retinopathy.

12, TITLE:      An Annotated Corpus of Emerging Anglicisms in Spanish Newspaper Headlines
http://arxiv.org/abs/2004.02929
AUTHORS:        Elena Álvarez-Mellado
COMMENTS:       Accepted at the 4th Workshop on Computational Approaches to Linguistic Code-Switching, LREC 2020
HIGHLIGHT:      In this paper we present: (1) a corpus of 21,570 newspaper headlines written in European Spanish annotated with emergent anglicisms and (2) a conditional random field baseline model with handcrafted features for anglicism extraction. We introduce a corpus of European Spanish newspaper headlines annotated with anglicisms and a baseline model for anglicism extraction. We present the newspaper headlines corpus, describe the annotation tagset and guidelines and introduce a CRF model that can serve as baseline for the task of detecting anglicisms.

13, TITLE:      Uniform State Abstraction For Reinforcement Learning
http://arxiv.org/abs/2004.02919
AUTHORS:        John Burden ; Daniel Kudenko
COMMENTS:       8 Pages, 2 figures, Accepted for publication in the European Conference of Artificial Intelligence (ECAI 2020)
HIGHLIGHT:      In this paper we extend and improve MRL to take advantage of modern Deep Learning algorithms such as Deep Q-Networks (DQN).

14, TITLE:      Speaker-change Aware CRF for Dialogue Act Classification
http://arxiv.org/abs/2004.02913
AUTHORS:        Guokan Shang ; Antoine Jean-Pierre Tixier ; Michalis Vazirgiannis ; Jean-Pierre Lorré
HIGHLIGHT:      To address this limitation, this paper proposes a simple modification of the CRF layer that takes speaker-change into account.

15, TITLE:      Fingerprint Presentation Attack Detection: A Sensor and Material Agnostic Approach
http://arxiv.org/abs/2004.02941
AUTHORS:        Steven A. Grosz ; Tarang Chugh ; Anil K. Jain
HIGHLIGHT:      In this study, we propose a robust PAD solution with improved cross-material and cross-sensor generalization.

16, TITLE:      Embedding Java Classes with code2vec: Improvements from Variable Obfuscation
http://arxiv.org/abs/2004.02942
AUTHORS:        Rhys Compton ; Eibe Frank ; Panos Patros ; Abigail Koay
COMMENTS:       In 17th International Conference on Mining Software Repositories (MSR) 2020, Seoul, Republic of Korea. 11 pages
HIGHLIGHT:      Both shortcomings are addressed in the research presented in this paper.

17, TITLE:      Objectness-Aware One-Shot Semantic Segmentation
http://arxiv.org/abs/2004.02945
AUTHORS:        Yinan Zhao ; Brian Price ; Scott Cohen ; Danna Gurari
HIGHLIGHT:      In this paper, we tackle the challenging one-shot semantic segmentation problem by taking advantage of objectness.

18, TITLE:      Beyond Background-Aware Correlation Filters: Adaptive Context Modeling by Hand-Crafted and Deep RGB Features for Visual Tracking
http://arxiv.org/abs/2004.02932
AUTHORS:        Seyed Mojtaba Marvasti-Zadeh ; Hossein Ghanei-Yakhdan ; Shohreh Kasaei
HIGHLIGHT:      In this paper, an adaptive background-aware correlation filter-based tracker is proposed that effectively models the target appearance by using either the histogram of oriented gradients (HOG) or convolutional neural network (CNN) feature maps.

19, TITLE:      Efficient Scale Estimation Methods using Lightweight Deep Convolutional Neural Networks for Visual Tracking
http://arxiv.org/abs/2004.02933
AUTHORS:        Seyed Mojtaba Marvasti-Zadeh ; Hossein Ghanei-Yakhdan ; Shohreh Kasaei
HIGHLIGHT:      Whereas the exploitation of CNNs imposes a high computational burden, this paper exploits pre-trained lightweight CNNs models to propose two efficient scale estimation methods, which not only improve the visual tracking performance but also provide acceptable tracking speeds.

20, TITLE:      Evolving Normalization-Activation Layers
http://arxiv.org/abs/2004.02967
AUTHORS:        Hanxiao Liu ; Andrew Brock ; Karen Simonyan ; Quoc V. Le
HIGHLIGHT:      Evolving Normalization-Activation Layers

21, TITLE:      Deblurring using Analysis-Synthesis Networks Pair
http://arxiv.org/abs/2004.02956
AUTHORS:        Adam Kaufman ; Raanan Fattal
HIGHLIGHT:      We propose a new architecture which breaks the deblurring network into an analysis network which estimates the blur, and a synthesis network that uses this kernel to deblur the image.

22, TITLE:      TSInsight: A local-global attribution framework for interpretability in time-series data
http://arxiv.org/abs/2004.02958
AUTHORS:        Shoaib Ahmed Siddiqui ; Dominique Mercier ; Andreas Dengel ; Sheraz Ahmed
HIGHLIGHT:      We approach the problem of interpretability in a novel way by proposing TSInsight where we attach an auto-encoder to the classifier with a sparsity-inducing norm on its output and fine-tune it based on the gradients from the classifier and a reconstruction penalty.

23, TITLE:      COVID-Xpert: An AI Powered Population Screening of COVID-19 Cases Using Chest Radiography Images
http://arxiv.org/abs/2004.03042
AUTHORS:        Xin Li ; Dongxiao Zhu
COMMENTS:       COVID-19 Population Screening, CoVid-19, X-ray Chest Radiography, Computer-Aided Diagnosis, Deep Learning, Artificial Intelligence
HIGHLIGHT:      We demonstrate the strong potential of our XCR based population screening approach, COVID-Xpert, for detecting COVID-19 cases through an impressive experimental performance.

24, TITLE:      When, Where, and What? A New Dataset for Anomaly Detection in Driving Videos
http://arxiv.org/abs/2004.03044
AUTHORS:        Yu Yao ; Xizi Wang ; Mingze Xu ; Zelin Pu ; Ella Atkins ; David Crandall
COMMENTS:       23 pages, 11 figures, 6 tables
HIGHLIGHT:      This paper proposes traffic anomaly detection with a \textit{when-where-what} pipeline to detect, localize, and recognize anomalous events from egocentric videos. We introduce a new dataset called Detection of Traffic Anomaly (DoTA) containing 4,677 videos with temporal, spatial, and categorical annotations.

25, TITLE:      A Corpus Study and Annotation Schema for Named Entity Recognition and Relation Extraction of Business Products
http://arxiv.org/abs/2004.03287
AUTHORS:        Saskia Schön ; Veselina Mironova ; Aleksandra Gabryszak ; Leonhard Hennig
COMMENTS:       Published in LREC 2018
HIGHLIGHT:      In this work, we present a corpus study, an annotation schema and associated guidelines, for the annotation of product entity and company-product relation mentions. We also describe our ongoing annotation effort, and present a preliminary corpus of English web and social media documents annotated according to the proposed guidelines.

26, TITLE:      Manifold-driven Attention Maps for Weakly Supervised Segmentation
http://arxiv.org/abs/2004.03046
AUTHORS:        Sukesh Adiga V ; Jose Dolz ; Herve Lombaert
COMMENTS:       Paper is submitted to MICCAI2020
HIGHLIGHT:      In this work, we propose a manifold driven attention-based network to enhance visual salient regions, thereby improving segmentation accuracy in a weakly supervised setting.

27, TITLE:      KorNLI and KorSTS: New Benchmark Datasets for Korean Natural Language Understanding
http://arxiv.org/abs/2004.03289
AUTHORS:        Jiyeon Ham ; Yo Joong Choe ; Kyubyong Park ; Ilji Choi ; Hyungjoon Soh
COMMENTS:       Access datasets at https://github.com/kakaobrain/KorNLUDatasets
HIGHLIGHT:      Motivated by this, we construct and release new datasets for Korean NLI and STS, dubbed KorNLI and KorSTS, respectively.

28, TITLE:      Depth Sensing Beyond LiDAR Range
http://arxiv.org/abs/2004.03048
AUTHORS:        Kai Zhang ; Jiaxin Xie ; Noah Snavely ; Qifeng Chen
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      To that end, we propose a novel three-camera system that utilizes small field of view cameras.

29, TITLE:      Teacher-Class Network: A Neural Network Compression Mechanism
http://arxiv.org/abs/2004.03281
AUTHORS:        Shaiq Munir Malik ; Mohbat Tharani ; Murtaza Taj
HIGHLIGHT:      In this paper, we propose a novel method called a teacher-class network consisting of a single teacher and multiple student networks (i.e. class of students).

30, TITLE:      A German Corpus for Fine-Grained Named Entity Recognition and Relation Extraction of Traffic and Industry Events
http://arxiv.org/abs/2004.03283
AUTHORS:        Martin Schiersch ; Veselina Mironova ; Maximilian Schmitt ; Philippe Thomas ; Aleksandra Gabryszak ; Leonhard Hennig
COMMENTS:       Published in LREC 2018
HIGHLIGHT:      This work describes a corpus of German-language documents which has been annotated with fine-grained geo-entities, such as streets, stops and routes, as well as standard named entity types.

31, TITLE:      Learning Generative Models of Shape Handles
http://arxiv.org/abs/2004.03028
AUTHORS:        Matheus Gadelha ; Giorgio Gori ; Duygu Ceylan ; Radomir Mech ; Nathan Carr ; Tamy Boubekeur ; Rui Wang ; Subhransu Maji
COMMENTS:       11 pages, 11 figures, accepted do CVPR 2020
HIGHLIGHT:      We present a generative model to synthesize 3D shapes as sets of handles -- lightweight proxies that approximate the original 3D shape -- for applications in interactive editing, shape parsing, and building compact 3D representations.

32, TITLE:      A Systematic Analysis of Morphological Content in BERT Models for Multiple Languages
http://arxiv.org/abs/2004.03032
AUTHORS:        Daniel Edmiston
HIGHLIGHT:      This work describes experiments which probe the hidden representations of several BERT-style models for morphological content.

33, TITLE:      The Role of Pragmatic and Discourse Context in Determining Argument Impact
http://arxiv.org/abs/2004.03034
AUTHORS:        Esin Durmus ; Faisal Ladhak ; Claire Cardie
COMMENTS:       EMNLP 2019
HIGHLIGHT:      This paper presents a new dataset to initiate the study of this aspect of argumentation: it consists of a diverse collection of arguments covering 741 controversial topics and comprising over 47,000 claims.

34, TITLE:      Directional approach to gradual cover: the continuous case
http://arxiv.org/abs/2004.03035
AUTHORS:        Tammy Drezner ; Zvi Drezner ; Pawel Kalczynski
HIGHLIGHT:      The objective of the cover location models is covering demand by facilities within a given distance.

35, TITLE:      Dense Steerable Filter CNNs for Exploiting Rotational Symmetry in Histology Images
http://arxiv.org/abs/2004.03037
AUTHORS:        Simon Graham ; David Epstein ; Nasir Rajpoot
HIGHLIGHT:      In this paper, we propose Dense Steerable Filter CNNs (DSF-CNNs) that use group convolutions with multiple rotated copies of each filter in a densely connected framework.

36, TITLE:      Autoencoders for Unsupervised Anomaly Segmentation in Brain MR Images: A Comparative Study
http://arxiv.org/abs/2004.03271
AUTHORS:        Christoph Baur ; Stefan Denner ; Benedikt Wiestler ; Shadi Albarqouni ; Nassir Navab
HIGHLIGHT:      The main principle behind these works is to learn a model of normal anatomy by learning to compress and recover healthy data.

37, TITLE:      Super-resolution of clinical CT volumes with modified CycleGAN using micro CT volumes
http://arxiv.org/abs/2004.03272
AUTHORS:        Tong ZHENG ; Hirohisa ODA ; Takayasu MORIYA ; Takaaki SUGINO ; Shota NAKAMURA ; Masahiro ODA ; Masaki MORI ; Hirotsugu TAKABATAKE ; Hiroshi NATORI ; Kensaku MORI
COMMENTS:       6 pages, 2 figures
HIGHLIGHT:      This paper presents a super-resolution (SR) method with unpaired training dataset of clinical CT and micro CT volumes.

38, TITLE:      MGGR: MultiModal-Guided Gaze Redirection with Coarse-to-Fine Learning
http://arxiv.org/abs/2004.03064
AUTHORS:        Jingjing Chen ; Jichao Zhang ; Jiayuan Fan ; Tao Chen ; Enver Sangineto ; Nicu Sebe
HIGHLIGHT:      To this end, we propose an innovative MultiModal-Guided Gaze Redirection~(MGGR) framework that fully exploits eye-map images and target angles to adjust a given eye appearance through a designed coarse-to-fine learning.

39, TITLE:      Are Natural Language Inference Models IMPPRESsive? Learning IMPlicature and PRESupposition
http://arxiv.org/abs/2004.03066
AUTHORS:        Paloma Jeretic ; Alex Warstadt ; Suvrat Bhooshan ; Adina Williams
HIGHLIGHT:      We create an IMPlicature and PRESupposition diagnostic dataset (IMPPRES), consisting of 32K semi-automatically generated sentence pairs illustrating well-studied pragmatic inference types.

40, TITLE:      Information-Theoretic Probing for Linguistic Structure
http://arxiv.org/abs/2004.03061
AUTHORS:        Tiago Pimentel ; Josef Valvoda ; Rowan Hall Maudslay ; Ran Zmigrod ; Adina Williams ; Ryan Cotterell
COMMENTS:       Accepted for publication at ACL 2020
HIGHLIGHT:      We propose an information-theoretic formalization of probing as estimating mutual information that contradicts this received wisdom: one should always select the highest performing probe one can, even if it is more complex, since it will result in a tighter estimate.

41, TITLE:      Scenario-Transferable Semantic Graph Reasoning for Interaction-Aware Probabilistic Prediction
http://arxiv.org/abs/2004.03053
AUTHORS:        Yeping Hu ; Wei Zhan ; Masayoshi Tomizuka
COMMENTS:       16 pages, 10 figures
HIGHLIGHT:      In this paper, we propose a scenario-transferable and interaction-aware probabilistic prediction algorithm based on semantic graph reasoning, which predicts behaviors of selected agents.

42, TITLE:      End-to-End Pseudo-LiDAR for Image-Based 3D Object Detection
http://arxiv.org/abs/2004.03080
AUTHORS:        Rui Qian ; Divyansh Garg ; Yan Wang ; Yurong You ; Serge Belongie ; Bharath Hariharan ; Mark Campbell ; Kilian Q. Weinberger ; Wei-Lun Chao
COMMENTS:       Accepted to 2020 Conference on Computer Vision and Pattern Recognition (CVPR 2020)
HIGHLIGHT:      In this paper, we introduce a new framework based on differentiable Change of Representation (CoR) modules that allow the entire PL pipeline to be trained end-to-end.

43, TITLE:      egg: Easy, Efficient, and Extensible E-graphs
http://arxiv.org/abs/2004.03082
AUTHORS:        Max Willsey ; Yisu Remy Wang ; Oliver Flatt ; Chandrakana Nandi ; Pavel Panchekha ; Zachary Tatlock
COMMENTS:       18 pages, 8 figures
HIGHLIGHT:      In this work, we expand on the idea of equality saturation and re-propose E-graphs as a solution to a diverse set of optimization problems, addressing issues identified by past work.

44, TITLE:      Inferential Text Generation with Multiple Knowledge Sources and Meta-Learning
http://arxiv.org/abs/2004.03070
AUTHORS:        Daya Guo ; Akari Asai ; Duyu Tang ; Nan Duan ; Ming Gong ; Linjun Shou ; Daxin Jiang ; Jian Yin ; Ming Zhou
HIGHLIGHT:      In this work, we use multiple knowledge sources as fuels for the model.

45, TITLE:      A Method for Curation of Web-Scraped Face Image Datasets
http://arxiv.org/abs/2004.03074
AUTHORS:        Kai Zhang ; Vítor Albiero ; Kevin W. Bowyer
COMMENTS:       This paper will appear at IWBF 2020
HIGHLIGHT:      We propose a semi-automated method, where the goal is to have a clean dataset for testing face recognition methods, with similar quality across men and women, to support comparison of accuracy across gender.

46, TITLE:      Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation
http://arxiv.org/abs/2004.03097
AUTHORS:        Bowen Wu ; Huan Zhang ; Mengyuan Li ; Zongsheng Wang ; Qihang Feng ; Junhong Huang ; Baoxun Wang
HIGHLIGHT:      In this paper, we propose a sentence representation approximating oriented distillation framework that can distill the pre-trained BERT into a simple LSTM based model without specifying tasks.

47, TITLE:      Interview: A Large-Scale Open-Source Corpus of Media Dialog
http://arxiv.org/abs/2004.03090
AUTHORS:        Bodhisattwa Prasad Majumder ; Shuyang Li ; Jianmo Ni ; Julian McAuley
HIGHLIGHT:      We introduce 'Interview': a large-scale (105K conversations) media dialog dataset collected from news interview transcripts.

48, TITLE:      Exemplar Auditing for Multi-Label Biomedical Text Classification
http://arxiv.org/abs/2004.03093
AUTHORS:        Allen Schmaltz ; Andrew Beam
COMMENTS:       22 pages, 8 tables
HIGHLIGHT:      In this work, we generalize a recently proposed zero-shot sequence labeling method, "binary labeling via a convolutional decomposition", to the case where the available document-level human labels are themselves relatively high-dimensional.

49, TITLE:      Is Graph Structure Necessary for Multi-hop Reasoning?
http://arxiv.org/abs/2004.03096
AUTHORS:        Nan Shao ; Yiming Cui ; Ting Liu ; Shijin Wang ; Guoping Hu
COMMENTS:       5 pages
HIGHLIGHT:      Recently, many works attempt to model texts as graph structure and introduce graph neural networks to deal with it on many NLP tasks.In this paper, we investigate whether graph structure is necessary for multi-hop reasoning tasks and what role it plays.

50, TITLE:      Homophone-based Label Smoothing in End-to-End Automatic Speech Recognition
http://arxiv.org/abs/2004.03437
AUTHORS:        Yi Zheng ; Xianjie Yang ; Xuyong Dang
HIGHLIGHT:      A new label smoothing method that makes use of prior knowledge of a language at human level, homophone, is proposed in this paper for automatic speech recognition (ASR).

51, TITLE:      Beer Organoleptic Optimisation: Utilising Swarm Intelligence and Evolutionary Computation Methods
http://arxiv.org/abs/2004.03438
AUTHORS:        Mohammad Majid al-Rifaie ; Marc Cavazza
HIGHLIGHT:      We investigate the problem by using three swarm intelligence and evolutionary computation techniques that enable brewers to map physico-chemical properties to target organoleptic properties to design a specific brew.

52, TITLE:      Utilising Prior Knowledge for Visual Navigation: Distil and Adapt
http://arxiv.org/abs/2004.03222
AUTHORS:        M. Mahdi Kazemi Moghaddam ; Qi Wu ; Ehsan Abbasnejad ; Javen Shi
COMMENTS:       Under review
HIGHLIGHT:      In this paper, we propose to use externally learned prior knowledge of object relations, which is integrated to our model via constructing a neural graph.

53, TITLE:      U-Net Using Stacked Dilated Convolutions for Medical Image Segmentation
http://arxiv.org/abs/2004.03466
AUTHORS:        Shuhang Wang ; Szu-Yeu Hu ; Eugene Cheah ; Xiaohong Wang ; Jingchao Wang ; Lei Chen ; Masoud Baikpour ; Arinc Ozturk ; Qian Li ; Shinn-Huey Chou ; Connie Lehman ; Viksit Kumar ; Anthony Samir
COMMENTS:       8 pages MICCAI
HIGHLIGHT:      This paper proposes a novel U-Net variant using stacked dilated convolutions for medical image segmentation (SDU-Net).

54, TITLE:      Specific Single- and Multi-Objective Evolutionary Algorithmsfor the Chance-Constrained Knapsack Problem
http://arxiv.org/abs/2004.03205
AUTHORS:        Yue Xie ; Aneta Neumann ; Frank Neumann
COMMENTS:       Accepted for oral presentation at GECCO 2020
HIGHLIGHT:      In this paper, consider problem-specific single-objective and multi-objective approaches for the problem.

55, TITLE:      Neural Image Inpainting Guided with Descriptive Text
http://arxiv.org/abs/2004.03212
AUTHORS:        Lisai Zhang ; Qingcai Chen ; Baotian Hu ; Shuoran Jiang
COMMENTS:       6 pages, 2 tables
HIGHLIGHT:      To acquire more semantically accurate inpainting images, this paper proposes a novel inpainting model named \textit{N}eural \textit{I}mage Inpainting \textit{G}uided with \textit{D}escriptive \textit{T}ext (NIGDT).

56, TITLE:      How Do You Act? An Empirical Study to Understand Behavior of Deep Reinforcement Learning Agents
http://arxiv.org/abs/2004.03237
AUTHORS:        Richard Meyes ; Moritz Schneider ; Tobias Meisen
COMMENTS:       16 pages, currently under review for publication for the ECMLPKDD 2020
HIGHLIGHT:      In this empirical study, we address this lack of transparency following an idea that is inspired by research in the field of neuroscience.

57, TITLE:      Variational Question-Answer Pair Generation for Machine Reading Comprehension
http://arxiv.org/abs/2004.03238
AUTHORS:        Kazutoshi Shinoda ; Akiko Aizawa
HIGHLIGHT:      We present a deep generative model of question-answer (QA) pairs for machine reading comprehension.

58, TITLE:      Automated Utterance Generation
http://arxiv.org/abs/2004.03484
AUTHORS:        Soham Parikh ; Quaizar Vohra ; Mitul Tiwari
COMMENTS:       IAAI-20, Emerging Application Track
HIGHLIGHT:      In this paper, we propose an utterance generation system which 1) uses extractive summarization to extract important sentences from the description, 2) uses multiple paraphrasing techniques to generate a diverse set of paraphrases of the title and summary sentences, and 3) selects good candidate paraphrases with the help of a novel candidate selection algorithm.

59, TITLE:      A Few Topical Tweets are Enough for Effective User-Level Stance Detection
http://arxiv.org/abs/2004.03485
AUTHORS:        Younes Samih ; Kareem Darwish
HIGHLIGHT:      In this paper, we tackle stance detection for such users using two approaches.

60, TITLE:      Bayesian aggregation improves traditional single image crop classification approaches
http://arxiv.org/abs/2004.03468
AUTHORS:        Ivan Matvienko ; Mikhail Gasanov ; Anna Petrovskaia ; Raghavendra Belur Jana ; Maria Pukalchik ; Ivan Oseledets
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      We present a comparison between the classical ML approaches and U-Net NN for classifying crops with a single satellite image.

61, TITLE:      Improving Fluency of Non-Autoregressive Machine Translation
http://arxiv.org/abs/2004.03227
AUTHORS:        Zdeněk Kasner ; Jindřich Libovický ; Jindřich Helcl
HIGHLIGHT:      We train models for three language pairs: German, Czech, and Romanian from and into English.

62, TITLE:      Motion-supervised Co-Part Segmentation
http://arxiv.org/abs/2004.03234
AUTHORS:        Aliaksandr Siarohin* ; Subhankar Roy* ; Stéphane Lathuilière ; Sergey Tulyakov ; Elisa Ricci ; Nicu Sebe
HIGHLIGHT:      To overcome this limitation, we propose a self-supervised deep learning method for co-part segmentation.

63, TITLE:      SC4D: A Sparse 4D Convolutional Network for Skeleton-Based Action Recognition
http://arxiv.org/abs/2004.03259
AUTHORS:        Lei Shi ; Yifan Zhang ; Jian Cheng ; Hanqing Lu
HIGHLIGHT:      In this paper, a new perspective is presented for skeleton-based action recognition.

64, TITLE:      Enhancing Review Comprehension with Domain-Specific Commonsense
http://arxiv.org/abs/2004.03020
AUTHORS:        Aaron Traylor ; Chen Chen ; Behzad Golshan ; Xiaolan Wang ; Yuliang Li ; Yoshihiko Suhara ; Jinfeng Li ; Cagatay Demiralp ; Wang-Chiew Tan
COMMENTS:       8 pages
HIGHLIGHT:      In this paper, we introduce xSense, an effective system for review comprehension using domain-specific commonsense knowledge bases (xSense KBs).

65, TITLE:      Inspector Gadget: A Data Programming-based Labeling System for Industrial Images
http://arxiv.org/abs/2004.03264
AUTHORS:        Geon Heo ; Yuji Roh ; Seonghyeon Hwang ; Dayun Lee ; Steven Euijong Whang
COMMENTS:       10 pages, 12 figures
HIGHLIGHT:      In this work, we expand the horizon of data programming by directly applying it to images without this conversion, which is a common scenario for industrial applications.

66, TITLE:      Field-Level Crop Type Classification with k Nearest Neighbors: A Baseline for a New Kenya Smallholder Dataset
http://arxiv.org/abs/2004.03023
AUTHORS:        Hannah Kerner ; Catherine Nakalembe ; Inbal Becker-Reshef
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      In this paper, we provide context for the new western Kenya dataset which was collected during an atypical 2019 main growing season and demonstrate classification accuracy up to 64% for maize and 70% for cassava using k Nearest Neighbors--a fast, interpretable, and scalable method that can serve as a baseline for future work.

67, TITLE:      Self-Adjusting Evolutionary Algorithms for Multimodal Optimization
http://arxiv.org/abs/2004.03266
AUTHORS:        Amirhossein Rajabi ; Carsten Witt
COMMENTS:       26 pages. Full version of a paper appearing at GECCO 2020
HIGHLIGHT:      Recent theoretical research has shown that self-adjusting and self-adaptive mechanisms can provably outperform static settings in evolutionary algorithms for binary search spaces.

68, TITLE:      Guided Dialog Policy Learning without Adversarial Learning in the Loop
http://arxiv.org/abs/2004.03267
AUTHORS:        Ziming Li ; Sungjin Lee ; Baolin Peng ; Jinchao Li ; Shahin Shayandeh ; Jianfeng Gao
COMMENTS:       10 pages
HIGHLIGHT:      In this work, we propose to decompose the previous adversarial training into two different steps.

69, TITLE:      Query Focused Multi-Document Summarization with Distant Supervision
http://arxiv.org/abs/2004.03027
AUTHORS:        Yumo Xu ; Mirella Lapata
COMMENTS:       11 pages, 3 figures
HIGHLIGHT:      In this work, we leverage distant supervision from question answering where various resources are available to more explicitly capture the relationship between queries and documents.

70, TITLE:      Hierarchical Opacity Propagation for Image Matting
http://arxiv.org/abs/2004.03249
AUTHORS:        Yaoyi Li ; Qingyao Xu ; Hongtao Lu
HIGHLIGHT:      To this end, this paper presents a hierarchical opacity propagation (HOP) matting method, where the opacity information is propagated in the neighborhood of each point at different semantic levels.

71, TITLE:      The multi-objective optimisation of breakwaters using evolutionary approach
http://arxiv.org/abs/2004.03010
AUTHORS:        Nikolay O. Nikitin ; Iana S. Polonskaia ; Anna V. Kalyuzhnaya ; Alexander V. Boukhanovsky
HIGHLIGHT:      In the paper, the multi-objective evolutionary approach for the breakwaters optimisation is proposed.

72, TITLE:      LLHD: A Multi-level Intermediate Representation for Hardware Description Languages
http://arxiv.org/abs/2004.03494
AUTHORS:        Fabian Schuiki ; Andreas Kurth ; Tobias Grosser ; Luca Benini
HIGHLIGHT:      To solve this problem, we propose the LLHD multi-level IR.

73, TITLE:      "You are grounded!": Latent Name Artifacts in Pre-trained Language Models
http://arxiv.org/abs/2004.03012
AUTHORS:        Vered Shwartz ; Rachel Rudinger ; Oyvind Tafjord
HIGHLIGHT:      We focus on artifacts associated with the representation of given names (e.g., Donald), which, depending on the corpus, may be associated with specific entities, as indicated by next token prediction (e.g., Trump).

74, TITLE:      Adaptive Fractional Dilated Convolution Network for Image Aesthetics Assessment
http://arxiv.org/abs/2004.03015
AUTHORS:        Qiuyu Chen ; Wei Zhang ; Ning Zhou ; Peng Lei ; Yi Xu ; Yu Zheng ; Jianping Fan
COMMENTS:       Accepted by CVPR 2020
HIGHLIGHT:      In this paper, an adaptive fractional dilated convolution (AFDC), which is aspect-ratio-embedded, composition-preserving and parameter-free, is developed to tackle this issue natively in convolutional kernel level.

75, TITLE:      What do Models Learn from Question Answering Datasets?
http://arxiv.org/abs/2004.03490
AUTHORS:        Priyanka Sen ; Amir Saffari
HIGHLIGHT:      In this paper, we investigate what models are really learning from QA datasets by evaluating BERT-based models across five popular QA datasets.

76, TITLE:      Pyramid Focusing Network for mutation prediction and classification in CT images
http://arxiv.org/abs/2004.03302
AUTHORS:        Xukun Zhang ; Wenxin Hu ; Wen Wu
HIGHLIGHT:      In this paper, we propose a pyramid focusing network (PFNet) for mutation prediction and classification based on CT images.

77, TITLE:      Deep Multi-Shot Network for modelling Appearance Similarity in Multi-Person Tracking applications
http://arxiv.org/abs/2004.03531
AUTHORS:        María J. Gómez-Silva
HIGHLIGHT:      In order to reduce these tracking errors, and even their propagation in further frames, this article presents a Deep Multi-Shot neural model for measuring the Degree of Appearance Similarity (MS-DoAS) between person observations.

78, TITLE:      Multi-Task Learning via Co-Attentive Sharing for Pedestrian Attribute Recognition
http://arxiv.org/abs/2004.03164
AUTHORS:        Haitian Zeng ; Haizhou Ai ; Zijie Zhuang ; Long Chen
COMMENTS:       2020 IEEE International Conference on Multimedia & Expo
HIGHLIGHT:      In this paper, we propose a novel Co-Attentive Sharing (CAS) module which extracts discriminative channels and spatial regions for more effective feature sharing in multi-task learning.

79, TITLE:      Trying AGAIN instead of Trying Longer: Prior Learning for Automatic Curriculum Learning
http://arxiv.org/abs/2004.03168
AUTHORS:        Rémy Portelas ; Katja Hofmann ; Pierre-Yves Oudeyer
COMMENTS:       Accepted to the ICLR 2020 workshop Beyond tabula rasa in RL (BeTR-RL)
HIGHLIGHT:      Besides demonstrating 50% improvements on average over the current state of the art, the objective of this work is to give a first example of a new research direction oriented towards refining ACL techniques over multiple learners, which we call Classroom Teaching.

80, TITLE:      Adaptive Multiscale Illumination-Invariant Feature Representation for Undersampled Face Recognition
http://arxiv.org/abs/2004.03153
AUTHORS:        Yang Zhang ; Changhui Hu ; Xiaobo Lu
HIGHLIGHT:      This paper presents an novel illumination-invariant feature representation approach used to eliminate the varying illumination affection in undersampled face recognition.

81, TITLE:      Real-time Classification from Short Event-Camera Streams using Input-filtering Neural ODEs
http://arxiv.org/abs/2004.03156
AUTHORS:        Giorgio Giannone ; Asha Anoosheh ; Alessio Quaglino ; Pierluca D'Oro ; Marco Gallieri ; Jonathan Masci
COMMENTS:       Submitted to ICML 2020
HIGHLIGHT:      In this work, we instead propose to directly use events from a DVS camera, a stream of intensity changes and their spatial coordinates.

82, TITLE:      Deep Attentive Generative Adversarial Network for Photo-Realistic Image De-Quantization
http://arxiv.org/abs/2004.03150
AUTHORS:        Yang Zhang ; Changhui Hu ; Xiaobo Lu
HIGHLIGHT:      This paper proposes DAGAN algorithm to perform super-resolution on image intensity resolution, which is orthogonal to the spatial resolution, realizing photo-realistic de-quantization via an end-to-end learning pattern.

83, TITLE:      Self-Induced Curriculum Learning in Neural Machine Translation
http://arxiv.org/abs/2004.03151
AUTHORS:        Dana Ruiter ; Cristina España-Bonet ; Josef van Genabith
COMMENTS:       13 pages, 7 images
HIGHLIGHT:      In this study we provide an in-depth analysis of the sampling choices the SS-NMT model takes during training.

84, TITLE:      More Data, More Relations, More Context and More Openness: A Review and Outlook for Relation Extraction
http://arxiv.org/abs/2004.03186
AUTHORS:        Xu Han ; Tianyu Gao ; Yankai Lin ; Hao Peng ; Yaoliang Yang ; Chaojun Xiao ; Zhiyuan Liu ; Peng Li ; Maosong Sun ; Jie Zhou
HIGHLIGHT:      In this paper, we look back at existing RE methods, analyze key challenges we are facing nowadays, and show promising directions towards more powerful RE.

85, TITLE:      Increasing the Inference and Learning Speed of Tsetlin Machines with Clause Indexing
http://arxiv.org/abs/2004.03188
AUTHORS:        Saeed Rahimi Gorji ; Ole-Christoffer Granmo ; Sondre Glimsdal ; Jonathan Edwards ; Morten Goodwin
COMMENTS:       14 pages, 8 figures
HIGHLIGHT:      In this paper, we exploit this hierarchical structure by introducing a novel algorithm that avoids evaluating the clauses exhaustively.

86, TITLE:      Towards Multimodal Simultaneous Neural Machine Translation
http://arxiv.org/abs/2004.03180
AUTHORS:        Aizhan Imankulova ; Masahiro Kaneko ; Tosho Hirasawa ; Mamoru Komachi
HIGHLIGHT:      To alleviate this shortage, we propose multimodal simultaneous neural machine translation (MSNMT) which leverages visual information as an additional modality.

87, TITLE:      Multilingual enrichment of disease biomedical ontologies
http://arxiv.org/abs/2004.03181
AUTHORS:        Léo Bouscarrat ; Antoine Bonnefoy ; Cécile Capponi ; Carlos Ramisch
HIGHLIGHT:      We study the possibility to use open-source knowledge bases to translate biomedical ontologies.

88, TITLE:      Machine Translation with Unsupervised Length-Constraints
http://arxiv.org/abs/2004.03176
AUTHORS:        Jan Niehues
COMMENTS:       8 pages
HIGHLIGHT:      In this paper, we explore one of these, the generation of constraint translation.

89, TITLE:      Iconify: Converting Photographs into Icons
http://arxiv.org/abs/2004.03179
AUTHORS:        Takuro Karamatsu ; Gibran Benitez-Garcia ; Keiji Yanai ; Seiichi Uchida
COMMENTS:       to appear at 2020 Joint Workshop on Multimedia Artworks Analysis and Attractiveness Computing in Multimedia (MMArt-ACM'20)
HIGHLIGHT:      In this paper, we tackle a challenging domain conversion task between photo and icon images.

90, TITLE:      Decidability and Synthesis of Abstract Inductive Invariants
http://arxiv.org/abs/2004.03170
AUTHORS:        Francesco Ranzato
HIGHLIGHT:      We consider here inductive invariants belonging to an abstract domain $A$ as defined in abstract interpretation, namely, ensuring the existence of the best approximation in $A$ of any system property.

91, TITLE:      Multi-Scale Aggregation Using Feature Pyramid Module for Text-Independent Speaker Verification
http://arxiv.org/abs/2004.03194
AUTHORS:        Youngmoon Jung ; Seongmin Kye ; Yeunju Choi ; Myunghun Jung ; Hoirin Kim
COMMENTS:       Submitted to Interspeech 2020
HIGHLIGHT:      This paper improves the MSA by using a feature pyramid module, which enhances speaker-discriminative information of features at multiple layers via a top-down pathway and lateral connections.

92, TITLE:      Transformers to Learn Hierarchical Contexts in Multiparty Dialogue for Span-based Question Answering
http://arxiv.org/abs/2004.03561
AUTHORS:        Changmao Li ; Jinho D. Choi
COMMENTS:       Accepted by ACL 2020
HIGHLIGHT:      We introduce a novel approach to transformers that learns hierarchical representations in multiparty dialogue.

93, TITLE:      Windowing Models for Abstractive Summarization of Long Texts
http://arxiv.org/abs/2004.03324
AUTHORS:        Leon Schüller ; Florian Wilhelm ; Nico Kreiling ; Goran Glavaš
HIGHLIGHT:      We propose windowing models for neural abstractive summarization of (arbitrarily) long texts.

94, TITLE:      Dense Regression Network for Video Grounding
http://arxiv.org/abs/2004.03545
AUTHORS:        Runhao Zeng ; Haoming Xu ; Wenbing Huang ; Peihao Chen ; Mingkui Tan ; Chuang Gan
COMMENTS:       CVPR 2020
HIGHLIGHT:      The key idea of this paper is to use the distances between the frame within the ground truth and the starting (ending) frame as dense supervisions to improve the video grounding accuracy.

95, TITLE:      Towards Efficient Unconstrained Palmprint Recognition via Deep Distillation Hashing
http://arxiv.org/abs/2004.03303
AUTHORS:        Huikai Shao ; Dexing Zhong ; Xuefeng Du
COMMENTS:       13 pages, 8 figures, to access database, see http://gr.xjtu.edu.cn/web/bell/resource
HIGHLIGHT:      In this paper, a new palmprint benchmark is established, which consists of more than 20,000 images collected by 5 brands of smart phones in an unconstrained manner.

96, TITLE:      Unsupervised Person Re-identification via Softened Similarity Learning
http://arxiv.org/abs/2004.03547
AUTHORS:        Yutian Lin ; Lingxi Xie ; Yu Wu ; Chenggang Yan ; Qi Tian
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      In this paper, we follow the iterative training mechanism but discard clustering, since it incurs loss from hard quantization, yet its only product, image-level similarity, can be easily replaced by pairwise computation and a softened classification task.

97, TITLE:      Temporal Pyramid Network for Action Recognition
http://arxiv.org/abs/2004.03548
AUTHORS:        Ceyuan Yang ; Yinghao Xu ; Jianping Shi ; Bo Dai ; Bolei Zhou
COMMENTS:       To appear in CVPR 2020
HIGHLIGHT:      In this work we propose a generic Temporal Pyramid Network (TPN) at the feature-level, which can be flexibly integrated into 2D or 3D backbone networks in a plug-and-play manner.

98, TITLE:      Fine-Grained Named Entity Typing over Distantly Supervised Data Based on Refined Representations
http://arxiv.org/abs/2004.03554
AUTHORS:        Muhammad Asif Ali ; Yifang Sun ; Bing Li ; Wei Wang
HIGHLIGHT:      For this, we propose an edge-weighted attentive graph convolution network that refines the noisy mention representations by attending over corpus-level contextual clues prior to the end classification.

99, TITLE:      Entity Linking via Dual and Cross-Attention Encoders
http://arxiv.org/abs/2004.03555
AUTHORS:        Oshin Agarwal ; Daniel M. Bikel
HIGHLIGHT:      In this work, we use this retrieval system solely for generating candidate entities.

100, TITLE:     Class-Agnostic Continual Learning of Alternating Languages and Domains
http://arxiv.org/abs/2004.03340
AUTHORS:        Germán Kruszewski ; Ionut-Teodor Sorodoc ; Tomas Mikolov
HIGHLIGHT:      In this work, we propose a benchmark based on language modelling in a multilingual and multidomain setting that prescinds of any explicit delimitation of training examples into distinct tasks, and propose metrics to study continual learning and catastrophic forgetting in this setting.

101, TITLE:     Resultants over principal Artinian rings
http://arxiv.org/abs/2004.03341
AUTHORS:        Claus Fieker ; Tommy Hofmann ; Carlo Sircana
HIGHLIGHT:      Here we present an algorithm to compute it over Artinian principal rings with a modified version of the Euclidean algorithm.

102, TITLE:     DiagNet: towards a generic, Internet-scale root cause analysis solution
http://arxiv.org/abs/2004.03343
AUTHORS:        Loïck Bonniot ; Christoph Neumann ; François Taïani
HIGHLIGHT:      In this paper, we explore how different machine learning techniques can be used for Internet-scale root cause analysis using measurements taken from end-user devices.

103, TITLE:     Knowledge Fusion and Semantic Knowledge Ranking for Open Domain Question Answering
http://arxiv.org/abs/2004.03101
AUTHORS:        Pratyay Banerjee ; Chitta Baral
COMMENTS:       9 pages. 4 figures, 4 tables, [WIP]
HIGHLIGHT:      In our work, we learn a semantic knowledge ranking model to re-rank knowledge retrieved through Lucene based information retrieval systems.

104, TITLE:     Generalized Label Enhancement with Sample Correlations
http://arxiv.org/abs/2004.03104
AUTHORS:        Qinghai Zheng ; Jihua Zhu ; Haoyu Tang ; Xinyuan Liu ; Zhongyu Li ; Huimin Lu
HIGHLIGHT:      To handle this problem, we propose two novel label enhancement methods, i.e., Label Enhancement with Sample Correlations (LESC) and generalized Label Enhancement with Sample Correlations (gLESC).

105, TITLE:     Feature Pyramid Grids
http://arxiv.org/abs/2004.03580
AUTHORS:        Kai Chen ; Yuhang Cao ; Chen Change Loy ; Dahua Lin ; Christoph Feichtenhofer
COMMENTS:       Technical report
HIGHLIGHT:      In this paper, we present Feature Pyramid Grids (FPG), a deep multi-pathway feature pyramid, that represents the feature scale-space as a regular grid of parallel bottom-up pathways which are fused by multi-directional lateral connections.

106, TITLE:     Cascaded Refinement Network for Point Cloud Completion
http://arxiv.org/abs/2004.03327
AUTHORS:        Xiaogang Wang ; Marcelo H Ang Jr ; Gim Hee Lee
COMMENTS:       To appear in CVPR2020
HIGHLIGHT:      To this end, we propose a cascaded refinement network together with a coarse-to-fine strategy to synthesize the detailed object shapes.

107, TITLE:     MedDialog: A Large-scale Medical Dialogue Dataset
http://arxiv.org/abs/2004.03329
AUTHORS:        Shu Chen ; Zeqian Ju ; Xiangyu Dong ; Hongchao Fang ; Sicheng Wang ; Yue Yang ; Jiaqi Zeng ; Ruisi Zhang ; Ruoyu Zhang ; Meng Zhou ; Penghui Zhu ; Pengtao Xie
HIGHLIGHT:      To facilitate the research and development of medical dialogue systems, we build a large-scale medical dialogue dataset -- MedDialog -- that contains 1.1 million conversations between patients and doctors and 4 million utterances.

108, TITLE:     Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation
http://arxiv.org/abs/2004.03572
AUTHORS:        Jiaming Sun ; Linghao Chen ; Yiming Xie ; Siyu Zhang ; Qinhong Jiang ; Xiaowei Zhou ; Hujun Bao
COMMENTS:       Accepted to CVPR 2020. Code is available at https://github.com/zju3dv/disprcnn
HIGHLIGHT:      In this paper, we propose a novel system named Disp R-CNN for 3D object detection from stereo images.

109, TITLE:     Neural Analogical Matching
http://arxiv.org/abs/2004.03573
AUTHORS:        Maxwell Crouse ; Constantine Nakos ; Ibrahim Abdelaziz ; Kenneth Forbus
HIGHLIGHT:      As part of the first steps towards such an integration, we introduce the Analogical Matching Network; a neural architecture that learns to produce analogies between structured, symbolic representations that are largely consistent with the principles of Structure-Mapping Theory.

110, TITLE:     Two-Stage Resampling for Convolutional Neural Network Training in the Imbalanced Colorectal Cancer Image Classification
http://arxiv.org/abs/2004.03332
AUTHORS:        Michał Koziarski
HIGHLIGHT:      Two-Stage Resampling for Convolutional Neural Network Training in the Imbalanced Colorectal Cancer Image Classification

111, TITLE:     Event Based, Near Eye Gaze Tracking Beyond 10,000Hz
http://arxiv.org/abs/2004.03577
AUTHORS:        Anastasios N. Angelopoulos ; Julien N. P. Martel ; Amit P. S. Kohli ; Jorg Conradt ; Gordon Wetzstein
COMMENTS:       Technical report
HIGHLIGHT:      Here, we propose a hybrid frame-event-based near-eye gaze tracking system offering update rates beyond 10,000 Hz with an accuracy that matches that of high-end desktop-mounted commercial eye trackers when evaluated in the same conditions.

112, TITLE:     A Sentence Cloze Dataset for Chinese Machine Reading Comprehension
http://arxiv.org/abs/2004.03116
AUTHORS:        Yiming Cui ; Ting Liu ; Ziqing Yang ; Zhipeng Chen ; Wentao Ma ; Wanxiang Che ; Shijin Wang ; Guoping Hu
COMMENTS:       6 pages
HIGHLIGHT:      To add diversity in this area, in this paper, we propose a new task called Sentence Cloze-style Machine Reading Comprehension (SC-MRC).

113, TITLE:     RYANSQL: Recursively Applying Sketch-based Slot Fillings for Complex Text-to-SQL in Cross-Domain Databases
http://arxiv.org/abs/2004.03125
AUTHORS:        DongHyun Choi ; Myeong Cheol Shin ; EungGyun Kim ; Dong Ryeol Shin
COMMENTS:       10 pages, 1 figure
HIGHLIGHT:      In this paper, we present a neural network approach called RYANSQL (Recursively Yielding Annotation Network for SQL) to solve complex Text-to-SQL tasks for cross-domain databases.

114, TITLE:     Generative Adversarial Zero-shot Learning via Knowledge Graphs
http://arxiv.org/abs/2004.03109
AUTHORS:        Yuxia Geng ; Jiaoyan Chen ; Zhuo Chen ; Zhiquan Ye ; Zonggang Yuan ; Yantao Jia ; Huajun Chen
COMMENTS:       under review
HIGHLIGHT:      In this paper, we introduce a new generative ZSL method named KG-GAN by incorporating rich semantics in a knowledge graph (KG) into GANs.

115, TITLE:     Inexpensive Domain Adaptation of Pretrained Language Models: A Case Study on Biomedical Named Entity Recognition
http://arxiv.org/abs/2004.03354
AUTHORS:        Nina Poerner ; Ulli Waltinger ; Hinrich Schütze
HIGHLIGHT:      We cover over 50% of the BioBERT-BERT F1 delta, at 5% of BioBERT's CO_2 footprint and 2% of its cloud compute cost.

116, TITLE:     Inclusive GAN: Improving Data and Minority Coverage in Generative Models
http://arxiv.org/abs/2004.03355
AUTHORS:        Ning Yu ; Ke Li ; Peng Zhou ; Jitendra Malik ; Larry Davis ; Mario Fritz
HIGHLIGHT:      We develop an extension that allows explicit control over the minority subgroups that the model should ensure to include, and validate its effectiveness at little compromise from the overall performance on the entire dataset.

117, TITLE:     Human Motion Transfer from Poses in the Wild
http://arxiv.org/abs/2004.03142
AUTHORS:        Jian Ren ; Menglei Chai ; Sergey Tulyakov ; Chen Fang ; Xiaohui Shen ; Jianchao Yang
HIGHLIGHT:      In this paper, we tackle the problem of human motion transfer, where we synthesize novel motion video for a target person that imitates the movement from a reference video. To further boost research on the topic, we build two human motion datasets.

118, TITLE:     Predicting Camera Viewpoint Improves Cross-dataset Generalization for 3D Human Pose Estimation
http://arxiv.org/abs/2004.03143
AUTHORS:        Zhe Wang ; Daeyun Shin ; Charless C. Fowlkes
COMMENTS:       http://wangzheallen.github.io/cross-dataset-generalization
HIGHLIGHT:      In this work we carry out a systematic study of the diversity and biases present in specific datasets and its effect on cross-dataset generalization across a compendium of 5 pose datasets.

119, TITLE:     Efficient Context and Schema Fusion Networks for Multi-Domain Dialogue State Tracking
http://arxiv.org/abs/2004.03386
AUTHORS:        Su Zhu ; Jieyu Li ; Lu Chen ; Kai Yu
COMMENTS:       13 pages, 3 figures
HIGHLIGHT:      In this paper, a novel context and schema fusion network is proposed to encode the dialogue context and schema graph by using internal and external attention mechanisms.

120, TITLE:     Exact separation of forbidden-set cuts associated with redundant parity checks of binary linear codes
http://arxiv.org/abs/2004.03387
AUTHORS:        Christian Puchert ; Andreas M. Tillmann
HIGHLIGHT:      In this note, we prove NP-hardness of this problem.

121, TITLE:     Plug-and-play ISTA converges with kernel denoisers
http://arxiv.org/abs/2004.03145
AUTHORS:        Ruturaj G. Gavaskar ; Kunal N. Chaudhury
COMMENTS:       5 pages, Accepted to IEEE Signal Processing Letters
HIGHLIGHT:      We prove that, under reasonable assumptions, fixed-point convergence of PnP-ISTA is indeed guaranteed for linear inverse problems such as deblurring, inpainting and superresolution (the assumptions are verifiable for inpainting).

122, TITLE:     Improving BPSO-based feature selection applied to offline WI handwritten signature verification through overfitting control
http://arxiv.org/abs/2004.03373
AUTHORS:        Victor L. F. Souza ; Adriano L. I. Oliveira ; Rafael M. O. Cruz ; Robert Sabourin
HIGHLIGHT:      This paper investigates the presence of overfitting when using Binary Particle Swarm Optimization (BPSO) to perform the feature selection in a context of Handwritten Signature Verification (HSV).

123, TITLE:     Toward Fine-grained Facial Expression Manipulation
http://arxiv.org/abs/2004.03132
AUTHORS:        Jun Ling ; Han Xue ; Li Song ; Shuhui Yang ; Rong Xie ; Xiao Gu
HIGHLIGHT:      In this study, we take these two objectives into consideration and propose a novel conditional GAN model.

124, TITLE:     Neutralizing Gender Bias in Word Embedding with Latent Disentanglement and Counterfactual Generation
http://arxiv.org/abs/2004.03133
AUTHORS:        Seungjae Shin ; Kyungwoo Song ; JoonHo Jang ; Hyemi Kim ; Weonyoung Joo ; Il-Chul Moon
COMMENTS:       10 pages
HIGHLIGHT:      Whereas the previous debiasing models project word embeddings into a linear subspace, we introduce a Latent Disentangling model with a siamese auto-encoder structure and a gradient reversal layer.

125, TITLE:     g2pM: A Neural Grapheme-to-Phoneme Conversion Package for MandarinChinese Based on a New Open Benchmark Dataset
http://arxiv.org/abs/2004.03136
AUTHORS:        Kyubyong Park ; Seanie Lee
COMMENTS:       will be submitted to Interspeech 2020
HIGHLIGHT:      Motivated by these, in this work, we introduce a new benchmark dataset that consists of 99,000+ sentences for Chinese polyphone disambiguation.

126, TITLE:     Unsupervised Neural Machine Translation with Indirect Supervision
http://arxiv.org/abs/2004.03137
AUTHORS:        Hongxiao Bai ; Mingxuan Wang ; Hai Zhao ; Lei Li
HIGHLIGHT:      In this work, we introduce a multilingual unsupervised NMT (\method) framework to leverage weakly supervised signals from high-resource language pairs to zero-resource translation directions.



==========Updates to Previous Papers==========
1, TITLE:       Review of Artificial Intelligence Techniques in Imaging Data Acquisition, Segmentation and Diagnosis for COVID-19
http://arxiv.org/abs/2004.02731
AUTHORS:        Feng Shi ; Jun Wang ; Jun Shi ; Ziyan Wu ; Qian Wang ; Zhenyu Tang ; Kelei He ; Yinghuan Shi ; Dinggang Shen
COMMENTS:       Added journal submission info
HIGHLIGHT:      In this review paper, we thus cover the entire pipeline of medical imaging and analysis techniques involved with COVID-19, including image acquisition, segmentation, diagnosis, and follow-up.

2, TITLE:       A Simple Baseline for Multi-Object Tracking
http://arxiv.org/abs/2004.01888
AUTHORS:        Yifu Zhang ; Chunyu Wang ; Xinggang Wang ; Wenjun Zeng ; Wenyu Liu
HIGHLIGHT:      In this work, we study the essential reasons behind the failure, and accordingly present a simple baseline to addresses the problems.

3, TITLE:       Evaluating Scalable Bayesian Deep Learning Methods for Robust Computer Vision
http://arxiv.org/abs/1906.01620
AUTHORS:        Fredrik K. Gustafsson ; Martin Danelljan ; Thomas B. Schön
COMMENTS:       CVPR Workshops 2020. Code is available at https://github.com/fregu856/evaluating_bdl
HIGHLIGHT:      We therefore accept this task and propose a comprehensive evaluation framework for scalable epistemic uncertainty estimation methods in deep learning.

4, TITLE:       Cars Can't Fly up in the Sky: Improving Urban-Scene Segmentation via Height-driven Attention Networks
http://arxiv.org/abs/2003.05128
AUTHORS:        Sungha Choi ; Joanne T. Kim ; Jaegul Choo
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      This paper exploits the intrinsic features of urban-scene images and proposes a general add-on module, called height-driven attention networks (HANet), for improving semantic segmentation for urban-scene images.

5, TITLE:       Adaptive Partial Scanning Transmission Electron Microscopy with Reinforcement Learning
http://arxiv.org/abs/2004.02786
AUTHORS:        Jeffrey M. Ede
COMMENTS:       13 pages, 4 figures. Appendix: 7 pages
HIGHLIGHT:      We have extended recurrent deterministic policy gradients to train deep LSTMs and differentiable neural computers to adaptively sample scan path segments.

6, TITLE:       Message complexity of population protocols
http://arxiv.org/abs/2003.09532
AUTHORS:        Talley Amir ; James Aspnes ; David Doty ; Mahsa Eftekhari ; Eric Severson
HIGHLIGHT:      The standard population protocol model assumes that when two agents interact, each observes the entire state of the other agent.

7, TITLE:       Instance-based Transfer Learning for Multilingual Deep Retrieval
http://arxiv.org/abs/1911.06111
AUTHORS:        Andrew O. Arnold ; William W. Cohen
HIGHLIGHT:      Perhaps the simplest type of multilingual transfer learning is instance-based transfer learning, in which data from the target language and the auxiliary languages are pooled, and a single model is learned from the pooled data.

8, TITLE:       Efficient Object Detection in Large Images using Deep Reinforcement Learning
http://arxiv.org/abs/1912.03966
AUTHORS:        Burak Uzkent ; Christopher Yeh ; Stefano Ermon
COMMENTS:       Published in WACV 2020
HIGHLIGHT:      To reduce the large computational and monetary cost associated with using high spatial resolution images, we propose a reinforcement learning agent that adaptively selects the spatial resolution of each image that is provided to the detector.

9, TITLE:       WaveTTS: Tacotron-based TTS with Joint Time-Frequency Domain Loss
http://arxiv.org/abs/2002.00417
AUTHORS:        Rui Liu ; Berrak Sisman ; Feilong Bao ; Guanglai Gao ; Haizhou Li
COMMENTS:       To appear at Odyssey 2020, Tokyo, Japan
HIGHLIGHT:      To address this problem, we propose a new training scheme for Tacotron-based TTS, referred to as WaveTTS, that has 2 loss functions: 1) time-domain loss, denoted as the waveform loss, that measures the distortion between the natural and generated waveform; and 2) frequency-domain loss, that measures the Mel-scale acoustic feature loss between the natural and generated acoustic features.

10, TITLE:      Temporally Distributed Networks for Fast Video Semantic Segmentation
http://arxiv.org/abs/2004.01800
AUTHORS:        Ping Hu ; Fabian Caba Heilbron ; Oliver Wang ; Zhe Lin ; Stan Sclaroff ; Federico Perazzi
COMMENTS:       [CVPR2020] Project: https://github.com/feinanshan/TDNet
HIGHLIGHT:      We present TDNet, a temporally distributed network designed for fast and accurate video semantic segmentation.

11, TITLE:      Scene Recomposition by Learning-based ICP
http://arxiv.org/abs/1812.05583
AUTHORS:        Hamid Izadinia ; Steven M. Seitz
COMMENTS:       To appear at CVPR 2020
HIGHLIGHT:      The proposed method is evaluated on real scenes datasets of SceneNN and ScanNet as well as synthetic scenes of SUNCG.

12, TITLE:      Symbiosis Promotes Fitness Improvements in the Game of Life
http://arxiv.org/abs/1908.07034
AUTHORS:        Peter D. Turney
COMMENTS:       Major changes to Sections 4.2 and 4.4. Minor changes throughout. Note that figures and tables appear at the end of the document
HIGHLIGHT:      We present a computational simulation of evolving entities that includes symbiosis with shifting levels of selection.

13, TITLE:      3DRegNet: A Deep Neural Network for 3D Point Registration
http://arxiv.org/abs/1904.01701
AUTHORS:        G. Dias Pais ; Srikumar Ramalingam ; Venu Madhav Govindu ; Jacinto C. Nascimento ; Rama Chellappa ; Pedro Miraldo
COMMENTS:       15 pages, 8 figures, 6 tables
HIGHLIGHT:      With regard to regression, we present two alternative approaches: (i) a Deep Neural Network (DNN) registration and (ii) a Procrustes approach using SVD to estimate the transformation.

14, TITLE:      Actions as Moving Points
http://arxiv.org/abs/2001.04608
AUTHORS:        Yixuan Li ; Zixu Wang ; Limin Wang ; Gangshan Wu
COMMENTS:       Technical Report
HIGHLIGHT:      In this paper, we present a conceptually simple, computationally efficient, and more precise action tubelet detection framework, termed as MovingCenter Detector (MOC-detector), by treating an action instance as a trajectory of moving points.

15, TITLE:      MUXConv: Information Multiplexing in Convolutional Neural Networks
http://arxiv.org/abs/2003.13880
AUTHORS:        Zhichao Lu ; Kalyanmoy Deb ; Vishnu Naresh Boddeti
COMMENTS:       CVPR 2020
HIGHLIGHT:      To overcome this limitation, we present MUXConv, a layer that is designed to increase the flow of information by progressively multiplexing channel and spatial information in the network, while mitigating computational complexity.

16, TITLE:      Multi-modality super-resolution loss for GAN-based super-resolution of clinical CT images using micro CT image database
http://arxiv.org/abs/1912.12838
AUTHORS:        Tong Zheng ; Hirohisa Oda ; Takayasu Moriya ; Shota Nakamura ; Masahiro Oda ; Masaki Mori ; Horitsugu Takabatake ; Hiroshi Natori ; Kensaku Mori
COMMENTS:       6 pages, 2 figures
HIGHLIGHT:      This paper newly introduces multi-modality loss function for GAN-based super-resolution that can maintain image structure and intensity on unpaired training dataset of clinical CT and micro CT volumes.

17, TITLE:      An Empirical Evaluation of Perturbation-based Defenses
http://arxiv.org/abs/2002.03080
AUTHORS:        Adam Dziedzic ; Sanjay Krishnan
HIGHLIGHT:      Based on these insights, we demonstrate a new robust model built on noise injection and adversarial training that achieves state-of-the-art robustness.

18, TITLE:      Automated Detection of Left Ventricle in Arterial Input Function Images for Inline Perfusion Mapping using Deep Learning: A study of 15,000 Patients
http://arxiv.org/abs/1910.07122
AUTHORS:        Hui Xue ; Ethan Tseng ; Kristopher D Knott ; Tushar Kotecha ; Louise Brown ; Sven Plein ; Marianna Fontana ; James C Moon ; Peter Kellman
COMMENTS:       Accepted by Magnetic Resonance in Medicine on March 30, 2020
HIGHLIGHT:      For this purpose, this study presents a robust AIF detection method using the convolutional neural net (CNN) model.

19, TITLE:      On the Security Risk of Cancelable Biometrics
http://arxiv.org/abs/1910.07770
AUTHORS:        Xingbo Dong ; Zhe Jin ; Andrew Beng Jin Teoh ; Massimo Tistarelli ; KokSheik Wong
COMMENTS:       Submit to PR
HIGHLIGHT:      In this paper, we analyzed the property of distance preservation in cancelable biometrics, and subsequently, a pre-image attack is launched to break the security of cancelable biometrics under the Kerckhoffs's assumption, where the cancelable biometrics algorithm and parameters are known to the attackers.

20, TITLE:      Establishing Strong Baselines for the New Decade: Sequence Tagging, Syntactic and Semantic Parsing with BERT
http://arxiv.org/abs/1908.04943
AUTHORS:        Han He ; Jinho D. Choi
COMMENTS:       To be published in The Thirty-Third International Flairs Conference proceedings
HIGHLIGHT:      This paper presents new state-of-the-art models for three tasks, part-of-speech tagging, syntactic parsing, and semantic parsing, using the cutting-edge contextualized embedding framework known as BERT.

21, TITLE:      ParSeNet: A Parametric Surface Fitting Network for 3D Point Clouds
http://arxiv.org/abs/2003.12181
AUTHORS:        Gopal Sharma ; Difan Liu ; Evangelos Kalogerakis ; Subhransu Maji ; Siddhartha Chaudhuri ; Radomír Měch
HIGHLIGHT:      We propose a novel, end-to-end trainable, deep network called ParSeNet that decomposes a 3D point cloud into parametric surface patches, including B-spline patches as well as basic geometric primitives.

22, TITLE:      Attentive One-Dimensional Heatmap Regression for Facial Landmark Detection and Tracking
http://arxiv.org/abs/2004.02108
AUTHORS:        Shi Yin ; Shangfei Wang ; Xiaoping Chen ; Enhong Chen
HIGHLIGHT:      To address this, we propose a novel attentive one-dimensional heatmap regression method for facial landmark localization.

23, TITLE:      Data Manipulation: Towards Effective Instance Learning for Neural Dialogue Generation via Learning to Augment and Reweight
http://arxiv.org/abs/2004.02594
AUTHORS:        Hengyi Cai ; Hongshen Chen ; Yonghao Song ; Cheng Zhang ; Xiaofang Zhao ; Dawei Yin
COMMENTS:       To appear at ACL 2020 (long paper)
HIGHLIGHT:      In this paper, we propose a data manipulation framework to proactively reshape the data distribution towards reliable samples by augmenting and highlighting effective learning samples as well as reducing the effect of inefficient samples simultaneously.

24, TITLE:      Privacy Preserving Gaze Estimation using Synthetic Images via a Randomized Encoding Based Framework
http://arxiv.org/abs/1911.07936
AUTHORS:        Efe Bozkir ; Ali Burak Ünal ; Mete Akgün ; Enkelejda Kasneci ; Nico Pfeifer
COMMENTS:       In Symposium on Eye Tracking Research and Applications (ETRA '20)
HIGHLIGHT:      To tackle this challenge, we employ a privacy-preserving framework based on randomized encoding to train a Support Vector Regression model using synthetic eye images privately to estimate the human gaze.

25, TITLE:      On the Role of Conceptualization in Commonsense Knowledge Graph Construction
http://arxiv.org/abs/2003.03239
AUTHORS:        Mutian He ; Yangqiu Song ; Kun Xu ; Dong Yu
COMMENTS:       11 pages, 4 figures
HIGHLIGHT:      To deal with the innumerable entities involved with commonsense in the real world, we introduce to CKG construction methods conceptualization, i.e., to view entities mentioned in text as instances of specific concepts or vice versa.

26, TITLE:      Bootstrapping a Crosslingual Semantic Parser
http://arxiv.org/abs/2004.02585
AUTHORS:        Tom Sherborne ; Yumo Xu ; Mirella Lapata
HIGHLIGHT:      In this work, we propose to adapt a semantic parser trained on a single language, such as English, to new languages and multiple domains with minimal annotation.

27, TITLE:      Negative Training for Neural Dialogue Response Generation
http://arxiv.org/abs/1903.02134
AUTHORS:        Tianxing He ; James Glass
HIGHLIGHT:      In this work, we propose a framework named "Negative Training" to minimize such behaviors.

28, TITLE:      Quantification of Tomographic Patterns associated with COVID-19 from Chest CT
http://arxiv.org/abs/2004.01279
AUTHORS:        Shikha Chaganti ; Abishek Balachandran ; Guillaume Chabin ; Stuart Cohen ; Thomas Flohr ; Bogdan Georgescu ; Philippe Grenier ; Sasa Grbic ; Siqi Liu ; François Mellot ; Nicolas Murray ; Savvas Nicolaou ; William Parker ; Thomas Re ; Pina Sanelli ; Alexander W. Sauter ; Zhoubing Xu ; Youngjin Yoo ; Valentin Ziebandt ; Dorin Comaniciu
HIGHLIGHT:      Purpose: To present a method that automatically detects and quantifies abnormal tomographic patterns commonly present in COVID-19, namely Ground Glass Opacities (GGO) and consolidations.

29, TITLE:      ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks
http://arxiv.org/abs/1910.03151
AUTHORS:        Qilong Wang ; Banggu Wu ; Pengfei Zhu ; Peihua Li ; Wangmeng Zuo ; Qinghua Hu
COMMENTS:       Accepted to CVPR 2020; Project Page: https://github.com/BangguWu/ECANet
HIGHLIGHT:      To overcome the paradox of performance and complexity trade-off, this paper proposes an Efficient Channel Attention (ECA) module, which only involves a handful of parameters while bringing clear performance gain.

30, TITLE:      A Discriminator Improves Unconditional Text Generation without Updating the Generator
http://arxiv.org/abs/2004.02135
AUTHORS:        Xingyuan Chen ; Ping Cai ; Peng Jin ; Hongjun Wang ; Xingyu Dai ; Jiajun Chen
HIGHLIGHT:      We propose a novel mechanism to improve a text generator with a discriminator, which is trained to estimate the probability that a sample comes from real or generated data.

31, TITLE:      Color inference from semantic labeling for person search in videos
http://arxiv.org/abs/1911.13114
AUTHORS:        Jules Simon ; Guillaume-Alexandre Bilodeau ; David Steele ; Harshad Mahadik
COMMENTS:       8 pages, 7 figures ICIAR 2020
HIGHLIGHT:      We propose an explainable model to generate semantic color labels for person search.

32, TITLE:      Reinforced Multi-task Approach for Multi-hop Question Generation
http://arxiv.org/abs/2004.02143
AUTHORS:        Deepak Gupta ; Hardik Chauhan ; Asif Ekbal ; Pushpak Bhattacharyya
HIGHLIGHT:      We employ multitask learning with the auxiliary task of answer-aware supporting fact prediction to guide the question generator.

33, TITLE:      DCT-Conv: Coding filters in convolutional networks with Discrete Cosine Transform
http://arxiv.org/abs/2001.08517
AUTHORS:        Karol Chęciński ; Paweł Wawrzyński
COMMENTS:       6 pages, 2 figures, submitted for IJCNN 2020
HIGHLIGHT:      In this paper, the trained parameters define a frequency spectrum which is transformed into convolutional filters with Inverse Discrete Cosine Transform (IDCT, the same is applied in decompression from JPEG).

34, TITLE:      Appraisal Theories for Emotion Classification in Text
http://arxiv.org/abs/2003.14155
AUTHORS:        Jan Hofmann ; Enrica Troiano ; Kai Sassenberg ; Roman Klinger
HIGHLIGHT:      With this paper, we propose to make such interpretations of events explicit, following theories of cognitive appraisal of events and show their potential for emotion classification when being encoded in classification models.

35, TITLE:      DSA: More Efficient Budgeted Pruning via Differentiable Sparsity Allocation
http://arxiv.org/abs/2004.02164
AUTHORS:        Xuefei Ning ; Tianchen Zhao ; Wenshuo Li ; Peng Lei ; Yu Wang ; Huazhong Yang
HIGHLIGHT:      In this paper, we propose Differentiable Sparsity Allocation (DSA), an efficient end-to-end budgeted pruning flow.

36, TITLE:      PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search
http://arxiv.org/abs/1907.05737
AUTHORS:        Yuhui Xu ; Lingxi Xie ; Xiaopeng Zhang ; Xin Chen ; Guo-Jun Qi ; Qi Tian ; Hongkai Xiong
COMMENTS:       Accepted by ICLR2020
HIGHLIGHT:      In this paper, we present a novel approach, namely, Partially-Connected DARTS, by sampling a small part of super-network to reduce the redundancy in exploring the network space, thereby performing a more efficient search without comprising the performance.

37, TITLE:      Defective Convolutional Networks
http://arxiv.org/abs/1911.08432
AUTHORS:        Tiange Luo ; Tianle Cai ; Mengxiao Zhang ; Siyu Chen ; Di He ; Liwei Wang
HIGHLIGHT:      To mitigate the threat of such adversarial attacks, we propose defective convolutional networks that make predictions relying less on textural information but more on shape information by properly integrating defective convolutional layers into standard CNNs.

38, 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.

39, TITLE:      Transforming Spectrum and Prosody for Emotional Voice Conversion with Non-Parallel Training Data
http://arxiv.org/abs/2002.00198
AUTHORS:        Kun Zhou ; Berrak Sisman ; Haizhou Li
COMMENTS:       accepted by Speaker Odyssey 2020 in Tokyo, Japan
HIGHLIGHT:      We propose a CycleGAN network to find an optimal pseudo pair from non-parallel training data by learning forward and inverse mappings simultaneously using adversarial and cycle-consistency losses.

40, TITLE:      Interpreting a Recurrent Neural Network's Predictions of ICU Mortality Risk
http://arxiv.org/abs/1905.09865
AUTHORS:        Long V. Ho ; Melissa D. Aczon ; David Ledbetter ; Randall Wetzel
HIGHLIGHT:      The goals of this work were to highlight which features contributed to a recurrent neural network's (RNN) predictions of ICU mortality and compare this information with clinical expectations.

41, TITLE:      Computational Complexity of the Hylland-Zeckhauser Scheme for One-Sided Matching Markets
http://arxiv.org/abs/2004.01348
AUTHORS:        Vijay V. Vazirani ; Mihalis Yannakakis
COMMENTS:       22 pages
HIGHLIGHT:      We present the following partial resolution: 1.

42, TITLE:      A Local-to-Global Approach to Multi-modal Movie Scene Segmentation
http://arxiv.org/abs/2004.02678
AUTHORS:        Anyi Rao ; Linning Xu ; Yu Xiong ; Guodong Xu ; Qingqiu Huang ; Bolei Zhou ; Dahua Lin
COMMENTS:       Accepted to CVPR2020. Project page: https://anyirao.com/projects/SceneSeg.html
HIGHLIGHT:      A Local-to-Global Approach to Multi-modal Movie Scene Segmentation

43, TITLE:      Work in Progress: Temporally Extended Auxiliary Tasks
http://arxiv.org/abs/2004.00600
AUTHORS:        Craig Sherstan ; Bilal Kartal ; Pablo Hernandez-Leal ; Matthew E. Taylor
COMMENTS:       Accepted for the Adaptive and Learning Agents (ALA) Workshop at AAMAS 2020
HIGHLIGHT:      The primary purpose of the work presented here is to investigate the impact that an auxiliary task's prediction timescale has on the agent's policy performance.

44, TITLE:      How Does Gender Balance In Training Data Affect Face Recognition Accuracy?
http://arxiv.org/abs/2002.02934
AUTHORS:        Vítor Albiero ; Kai Zhang ; Kevin W. Bowyer
HIGHLIGHT:      This work investigates female under-representation in the training data is truly the cause of lower accuracy for females on test data.

45, TITLE:      BlackBox Toolkit: Intelligent Assistance to UI Design
http://arxiv.org/abs/2004.01949
AUTHORS:        Vinoth Pandian Sermuga Pandian ; Sarah Suleri
COMMENTS:       Workshop position paper for CHI'20, Workshop on Artificial Intelligence for HCI: A Modern Approach; 4 pages, 3 figures
HIGHLIGHT:      In this research, we propose to modify the UI design process by assisting it with artificial intelligence (AI).

46, TITLE:      3D Dynamic Point Cloud Denoising via Spatial-Temporal Graph Learning
http://arxiv.org/abs/2003.08355
AUTHORS:        Wei Hu ; Qianjiang Hu ; Zehua Wang ; Xiang Gao
COMMENTS:       It's a new version of '3D Dynamic Point Cloud Denoising via Spatio-temporal Graph Modeling' (1904.12284). It was mis-posted. We will submit a replacement of this version to 1904.12284
HIGHLIGHT:      In this paper, we represent dynamic point clouds naturally on graphs and address the denoising problem by inferring the underlying graph via spatio-temporal graph learning, exploiting both the intra-frame similarity and inter-frame consistency.

47, TITLE:      Training neural networks to mimic the brain improves object recognition performance
http://arxiv.org/abs/1905.10679
AUTHORS:        Callie Federer ; Haoyan Xu ; Alona Fyshe ; Joel Zylberberg
HIGHLIGHT:      Training neural networks to mimic the brain improves object recognition performance

48, TITLE:      DAPAS : Denoising Autoencoder to Prevent Adversarial attack in Semantic Segmentation
http://arxiv.org/abs/1908.05195
AUTHORS:        Seungju Cho ; Tae Joon Jun ; Byungsoo Oh ; Daeyoung Kim
COMMENTS:       Accepted to be published in: 2020 International Joint Conference on Neural Networks (IJCNN), Glasgow, July 19--24, 2020
HIGHLIGHT:      Nowadays, Deep learning techniques show dramatic performance on computer vision area, and they even outperform human.

49, TITLE:      Improved Code Summarization via a Graph Neural Network
http://arxiv.org/abs/2004.02843
AUTHORS:        Alexander LeClair ; Sakib Haque ; Lingfei Wu ; Collin McMillan
COMMENTS:       10 pages
HIGHLIGHT:      Therefore, in this paper, we present an approach that uses a graph-based neural architecture that better matches the default structure of the AST to generate these summaries.

50, TITLE:      Regularizing Class-wise Predictions via Self-knowledge Distillation
http://arxiv.org/abs/2003.13964
AUTHORS:        Sukmin Yun ; Jongjin Park ; Kimin Lee ; Jinwoo Shin
COMMENTS:       Accepted to CVPR 2020. Code is available at https://github.com/alinlab/cs-kd
HIGHLIGHT:      To mitigate the issue, we propose a new regularization method that penalizes the predictive distribution between similar samples.

51, TITLE:      Selective Attention Based Graph Convolutional Networks for Aspect-Level Sentiment Classification
http://arxiv.org/abs/1910.10857
AUTHORS:        Xiaochen Hou ; Jing Huang ; Guangtao Wang ; Kevin Huang ; Xiaodong He ; Bowen Zhou
HIGHLIGHT:      In this paper, we propose to employ graph convolutional networks (GCNs) on the dependency tree to learn syntax-aware representations of aspect terms.

52, TITLE:      Cross-Shape Graph Convolutional Networks
http://arxiv.org/abs/2003.09053
AUTHORS:        Dmitry Petrov ; Evangelos Kalogerakis
HIGHLIGHT:      We present a method that processes 3D point clouds by performing graph convolution operations across shapes.

53, TITLE:      ReADS: A Rectified Attentional Double Supervised Network for Scene Text Recognition
http://arxiv.org/abs/2004.02070
AUTHORS:        Qi Song ; Qianyi Jiang ; Nan Li ; Rui Zhang ; Xiaolin Wei
COMMENTS:       8 pages, 3 figures
HIGHLIGHT:      In this paper, we elaborately design a Rectified Attentional Double Supervised Network (ReADS) for general scene text recognition.

54, TITLE:      Application of Genetic Algorithm for More Efficient Multi-Layer Thickness Optimization in Solar Cells
http://arxiv.org/abs/1909.06447
AUTHORS:        Premkumar Vincent ; Gwenaelle Cunha Sergio ; Jaewon Jang ; In Man Kang ; Jaehoon Park ; Hyeok Kim ; Minho Lee ; Jin-Hyuk Bae
COMMENTS:       Published (https://www.mdpi.com/1996-1073/13/7/1726), Code (https://github.com/gcunhase/GeneticAlgorithm-SolarCells)
HIGHLIGHT:      We have used a well-studied P3HT-based structure to test our algorithm.

55, TITLE:      A Computer-Aided Diagnosis System Using Artificial Intelligence for Hip Fractures Significantly Improves the Diagnostic Rate of Residents. -Multi-Institutional Joint Development Research-
http://arxiv.org/abs/2003.12443
AUTHORS:        Yoichi Sato M. D. ; Yasuhiko Takegami M. D. ; Ph. D. ; Takamune Asamoto M. D. ; Yutaro Ono M. D. ; Tsugeno Hidetoshi M. D. ; Ryosuke Goto ; Akira Kitamura M. S. in IT ; Seiwa Honda B. L
COMMENTS:       6 pages, 3 tables, 7 figures. / author's homepage : https://www.fracture-ai.org
HIGHLIGHT:      We used the EffectiventNet-B4 framework with Pytorch 1.3 and Fast.ai.

56, TITLE:      The role of formalism in system requirements
http://arxiv.org/abs/1911.02564
AUTHORS:        Jean-Michel Bruel ; Sophie Ebersold ; Florian Galinier ; Alexandr Naumchev ; Manuel Mazzara ; Bertrand Meyer
COMMENTS:       Third version (7 April 2020)
HIGHLIGHT:      The review includes examples from all of these categories, altogether 22 different methods, including for example SysML, Relax, Eiffel, Event-B, Alloy.

57, TITLE:      Cumulant-free closed-form formulas for some common (dis)similarities between densities of an exponential family
http://arxiv.org/abs/2003.02469
AUTHORS:        Frank Nielsen ; Richard Nock
COMMENTS:       33 pages
HIGHLIGHT:      In this work, we report (dis)similarity formulas which bypass the explicit use of the cumulant function and highlight the role of quasi-arithmetic means and their multivariate mean operator extensions.

58, TITLE:      How Additional Knowledge can Improve Natural Language Commonsense Question Answering?
http://arxiv.org/abs/1909.08855
AUTHORS:        Arindam Mitra ; Pratyay Banerjee ; Kuntal Kumar Pal ; Swaroop Mishra ; Chitta Baral
COMMENTS:       14 pages, 14 figures, 3 tables
HIGHLIGHT:      Our goal in this work is to develop methods to incorporate additional (commonsense) knowledge into language model-based approaches for better question-answering in such domains.

59, TITLE:      SketchyCOCO: Image Generation from Freehand Scene Sketches
http://arxiv.org/abs/2003.02683
AUTHORS:        Chengying Gao ; Qi Liu ; Qi Xu ; Limin Wang ; Jianzhuang Liu ; Changqing Zou
HIGHLIGHT:      We introduce the first method for automatic image generation from scene-level freehand sketches. We have built a large-scale composite dataset called SketchyCOCO to support and evaluate the solution.

60, TITLE:      Knowledge Guided Named Entity Recognition for BioMedical Text
http://arxiv.org/abs/1911.03869
AUTHORS:        Pratyay Banerjee ; Kuntal Kumar Pal ; Murthy Devarakonda ; Chitta Baral
COMMENTS:       6 pages, 2 figures, 5 tables, WIP
HIGHLIGHT:      In this work, we formulate the NER task as a multi-answer question answering (MAQA) task and provide different knowledge contexts, such as entity types, questions, definitions and definitions with examples.

61, TITLE:      Evolutionary Multi-Objective Optimization Driven by Generative Adversarial Networks
http://arxiv.org/abs/1907.04482
AUTHORS:        Cheng He ; Shihua Huang ; Ran Cheng ; Kay Chen Tan ; Yaochu Jin
COMMENTS:       This is a redundant version of 1910.04966
HIGHLIGHT:      Recently, more and more works have proposed to drive evolutionary algorithms using machine learning models.Usually, the performance of such model based evolutionary algorithms is highly dependent on the training qualities of the adopted models.Since it usually requires a certain amount of data (i.e. the candidate solutions generated by the algorithms) for model training, the performance deteriorates rapidly with the increase of the problem scales, due to the curse of dimensionality.To address this issue, we propose a multi-objective evolutionary algorithm driven by the generative adversarial networks (GANs).

62, TITLE:      A Generalized Multi-Task Learning Approach to Stereo DSM Filtering in Urban Areas
http://arxiv.org/abs/2004.02493
AUTHORS:        Lukas Liebel ; Ksenia Bittner ; Marco Körner
COMMENTS:       This paper was accepted for publication in the ISPRS Journal of Photogrammetry and Remote Sensing
HIGHLIGHT:      We propose a modular multi-task learning concept that consolidates existing approaches into a generalized framework.

63, TITLE:      Learning to Deceive with Attention-Based Explanations
http://arxiv.org/abs/1909.07913
AUTHORS:        Danish Pruthi ; Mansi Gupta ; Bhuwan Dhingra ; Graham Neubig ; Zachary C. Lipton
COMMENTS:       Accepted to ACL 2020 as a long paper. Updated version
HIGHLIGHT:      We call the latter use of attention mechanisms into question by demonstrating a simple method for training models to produce deceptive attention masks.

64, TITLE:      DOPS: Learning to Detect 3D Objects and Predict their 3D Shapes
http://arxiv.org/abs/2004.01170
AUTHORS:        Mahyar Najibi ; Guangda Lai ; Abhijit Kundu ; Zhichao Lu ; Vivek Rathod ; Thomas Funkhouser ; Caroline Pantofaru ; David Ross ; Larry S. Davis ; Alireza Fathi
COMMENTS:       To appear in CVPR 2020
HIGHLIGHT:      We propose DOPS, a fast single-stage 3D object detection method for LIDAR data.

65, TITLE:      CoPhy: Counterfactual Learning of Physical Dynamics
http://arxiv.org/abs/1909.12000
AUTHORS:        Fabien Baradel ; Natalia Neverova ; Julien Mille ; Greg Mori ; Christian Wolf
COMMENTS:       ICLR 2020 -Spotlight presentation
HIGHLIGHT:      We develop the CoPhy benchmark to assess the capacity of the state-of-the-art models for causal physical reasoning in a synthetic 3D environment and propose a model for learning the physical dynamics in a counterfactual setting. This work poses a new problem of counterfactual learning of object mechanics from visual input.

66, TITLE:      Learning ergodic averages in chaotic systems
http://arxiv.org/abs/2001.04027
AUTHORS:        Francisco Huhn ; Luca Magri
COMMENTS:       9 pages, 4 figures
HIGHLIGHT:      We propose a physics-informed machine learning method to predict the time average of a chaotic attractor.

67, TITLE:      The Knowledge Within: Methods for Data-Free Model Compression
http://arxiv.org/abs/1912.01274
AUTHORS:        Matan Haroush ; Itay Hubara ; Elad Hoffer ; Daniel Soudry
HIGHLIGHT:      We present three methods for generating synthetic samples from trained models.