2020.06.12.txt

==========New Papers==========
1, TITLE:       COVID-19-CT-CXR: a freely accessible and weakly labeled chest X-ray and CT image collection on COVID-19 from biomedical literature
http://arxiv.org/abs/2006.06177
AUTHORS:        Yifan Peng ; Yu-Xing Tang ; Sungwon Lee ; Yingying Zhu ; Ronald M. Summers ; Zhiyong Lu
HIGHLIGHT:      We extracted figures, associated captions, and relevant figure descriptions in the article and separated compound figures into subfigures.

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

3, TITLE:       Interpreting CNN for Low Complexity Learned Sub-pixel Motion Compensation in Video Coding
http://arxiv.org/abs/2006.06392
AUTHORS:        Luka Murn ; Saverio Blasi ; Alan F. Smeaton ; Noel E. O'Connor ; Marta Mrak
COMMENTS:       27th IEEE International Conference on Image Processing, 25-28 Oct 2020, Abu Dhabi, United Arab Emirates
HIGHLIGHT:      In this paper, a novel neural network-based tool is presented which improves the interpolation of reference samples needed for fractional precision motion compensation.

4, TITLE:       MOMS with Events: Multi-Object Motion Segmentation With Monocular Event Cameras
http://arxiv.org/abs/2006.06158
AUTHORS:        Chethan M. Parameshwara ; Nitin J. Sanket ; Arjun Gupta ; Cornelia Fermuller ; Yiannis Aloimonos
COMMENTS:       15 pages, 4 figures, Under review
HIGHLIGHT:      We propose a solution to multi-object motion segmentation using a combination of classical optimization methods along with deep learning and does not require prior knowledge of the 3D motion and the number and structure of objects.

5, TITLE:       Image Deconvolution via Noise-Tolerant Self-Supervised Inversion
http://arxiv.org/abs/2006.06156
AUTHORS:        Hirofumi Kobayashi ; Ahmet Can Solak ; Joshua Batson ; Loic A. Royer
HIGHLIGHT:      We propose a general framework for solving inverse problems in the presence of noise that requires no signal prior, no noise estimate, and no clean training data.

6, TITLE:       Enabling Nonlinear Manifold Projection Reduced-Order Models by Extending Convolutional Neural Networks to Unstructured Data
http://arxiv.org/abs/2006.06154
AUTHORS:        John Tencer ; Kevin Potter
COMMENTS:       Preprint
HIGHLIGHT:      We propose a nonlinear manifold learning technique based on deep autoencoders that is appropriate for model order reduction of physical systems in complex geometries. We propose sets of convolution operators based on the spatial derivative operators for the underlying spatial discretization, making the method particularly well suited to data arising from the solution of partial differential equations.

7, TITLE:       Emora STDM: A Versatile Framework for Innovative Dialogue System Development
http://arxiv.org/abs/2006.06143
AUTHORS:        James D. Finch ; Jinho D. Choi
COMMENTS:       Accepted by SIGDIAL 2020: System Demonstrations
HIGHLIGHT:      This demo paper presents Emora STDM (State Transition Dialogue Manager), a dialogue system development framework that provides novel workflows for rapid prototyping of chat-based dialogue managers as well as collaborative development of complex interactions.

8, TITLE:       TensorFlow with user friendly Graphical Framework for object detection API
http://arxiv.org/abs/2006.06385
AUTHORS:        Heemoon Yoon ; Sang-Hee Lee ; Mira Park
COMMENTS:       "The code of TF-GraF for TensorFlow object detection API is opened at https://github.com/boguss1225/ObjectDetectionGUI"
HIGHLIGHT:      Therefore, this is aim to develop an user friendly Graphical Framework for object detection API on TensorFlow which is called TensorFlow Graphical Framework (TF-GraF).

9, TITLE:       Kalman Filter Based Multiple Person Head Tracking
http://arxiv.org/abs/2006.06134
AUTHORS:        Mohib Ullah ; Maqsood Mahmud ; Habib Ullah ; Kashif Ahmad ; Ali Shariq Imran ; Faouzi Alaya Cheikh
COMMENTS:       5 pages, 2 figures
HIGHLIGHT:      In this paper, we come up with a simple yet effective target representation for human tracking.

10, TITLE:      Dance Revolution: Long Sequence Dance Generation with Music via Curriculum Learning
http://arxiv.org/abs/2006.06119
AUTHORS:        Ruozi Huang ; Huang Hu ; Wei Wu ; Kei Sawada ; Mi Zhang
COMMENTS:       Submitted to NeurIPS 2020
HIGHLIGHT:      To further alleviate the error accumulation in human motion synthesis, we introduce a dynamic auto-condition training strategy as a new curriculum learning method to facilitate the long-term dance generation.

11, TITLE:      Continual Learning for Affective Computing
http://arxiv.org/abs/2006.06113
AUTHORS:        Nikhil Churamani
COMMENTS:       Accepted at the Doctoral Consortium for the IEEE International Conference on Automatic Face and Gesture Recognition (FG), 2020
HIGHLIGHT:      In this dissertation, we propose the use of continual learning for affective computing as a paradigm for developing personalised affect perception.

12, TITLE:      Consolidating Commonsense Knowledge
http://arxiv.org/abs/2006.06114
AUTHORS:        Filip Ilievski ; Pedro Szekely ; Jingwei Cheng ; Fu Zhang ; Ehsan Qasemi
COMMENTS:       14 pages
HIGHLIGHT:      In this paper, we list representative sources and their properties.

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

14, TITLE:      Towards Unified Dialogue System Evaluation: A Comprehensive Analysis of Current Evaluation Protocols
http://arxiv.org/abs/2006.06110
AUTHORS:        Sarah E. Finch ; Jinho D. Choi
COMMENTS:       Accepted by SIGDIAL 2020
HIGHLIGHT:      This paper presents a comprehensive synthesis of both automated and human evaluation methods on dialogue systems, identifying their shortcomings while accumulating evidence towards the most effective evaluation dimensions.

15, TITLE:      PeopleMap: Visualization Tool for Mapping Out Researchers using Natural Language Processing
http://arxiv.org/abs/2006.06105
AUTHORS:        Jon Saad-Falcon ; Omar Shaikh ; Zijie J. Wang ; Austin P. Wright ; Sasha Richardson ; Duen Horng Chau
COMMENTS:       7 pages, 3 figures, submission to the 29th ACM International Conference on Information and Knowledge Management (CIKM '20), October 19-23, 2020, Galway, Ireland
HIGHLIGHT:      To solve this problem, we have developed PeopleMap, the first interactive, open-source, web-based tool that visually "maps out" researchers based on their research interests and publications by leveraging embeddings generated by natural language processing (NLP) techniques.

16, TITLE:      Towards Dynamic Algorithm Selection for Numerical Black-Box Optimization: Investigating BBOB as a Use Case
http://arxiv.org/abs/2006.06586
AUTHORS:        Diederick Vermetten ; Hao Wang ; Carola Doerr ; Thomas Bäck
HIGHLIGHT:      With this work we aim at promoting research on dynAC, by introducing a simpler variant that focuses only on switching between different algorithms, not configurations.

17, TITLE:      A framework for step-wise explaining how to solve constraint satisfaction problems
http://arxiv.org/abs/2006.06343
AUTHORS:        Bart Bogaerts ; Emilio Gamba ; Tias Guns
HIGHLIGHT:      Thereby, we aim to give the constraint solver explainable agency, which can help in building trust in the solver by being able to understand and even learn from the explanations.

18, TITLE:      Provenance for Linguistic Corpora Through Nanopublications
http://arxiv.org/abs/2006.06341
AUTHORS:        Timo Lek ; Anna de Groot ; Tobias Kuhn ; Roser Morante
HIGHLIGHT:      This paper addresses this issue with a case study on event annotated corpora and by creating a new, more interoperable representation of this data in the form of nanopublications.

19, TITLE:      Large-Scale Adversarial Training for Vision-and-Language Representation Learning
http://arxiv.org/abs/2006.06195
AUTHORS:        Zhe Gan ; Yen-Chun Chen ; Linjie Li ; Chen Zhu ; Yu Cheng ; Jingjing Liu
HIGHLIGHT:      We present VILLA, the first known effort on large-scale adversarial training for vision-and-language (V+L) representation learning.

20, TITLE:      An Edge Information and Mask Shrinking Based Image Inpainting Approach
http://arxiv.org/abs/2006.06196
AUTHORS:        Huali Xu ; Xiangdong Su ; Meng Wang ; Xiang Hao ; Guanglai Gao
COMMENTS:       Accepted by ICME2020
HIGHLIGHT:      To solve this problem, this paper proposes edge information and mask shrinking based image inpainting approach, which consists of two models.

21, TITLE:      Sensorimotor Visual Perception on Embodied System Using Free Energy Principle
http://arxiv.org/abs/2006.06192
AUTHORS:        Kanako Esaki ; Tadayuki Matsumura ; Kiyoto Ito ; Hiroyuki Mizuno
COMMENTS:       8 pages, 6 figures
HIGHLIGHT:      We propose an embodied system based on the free energy principle (FEP) for sensorimotor visual perception.

22, TITLE:      JIT-Masker: Efficient Online Distillation for Background Matting
http://arxiv.org/abs/2006.06185
AUTHORS:        Jo Chuang ; Qian Dong
HIGHLIGHT:      We design a real-time portrait matting pipeline for everyday use, particularly for "virtual backgrounds" in video conferences. We construct our own dataset of simulated video calls in various scenarios, and show that our approach delivers a 5x speedup over a saliency detection based pipeline in a non-GPU accelerated setting while delivering higher quality results.

23, TITLE:      G5: A Universal GRAPH-BERT for Graph-to-Graph Transfer and Apocalypse Learning
http://arxiv.org/abs/2006.06183
AUTHORS:        Jiawei Zhang
COMMENTS:       Keywords: Graph-Bert; Representation Learning; Apocalypse Learning; Transfer Learning; Graph Mining; Data Mining
HIGHLIGHT:      Two different label reasoning strategies, i.e., Cross-Source Classification Consistency Maximization (CCCM) and Cross-Source Dynamic Routing (CDR), are introduced in this paper to address the problem.

24, TITLE:      Transferring and Regularizing Prediction for Semantic Segmentation
http://arxiv.org/abs/2006.06570
AUTHORS:        Yiheng Zhang ; Zhaofan Qiu ; Ting Yao ; Chong-Wah Ngo ; Dong Liu ; Tao Mei
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we novelly exploit the intrinsic properties of semantic segmentation to alleviate such problem for model transfer.

25, TITLE:      Extracting and categorising the reactions to COVID-19 by the South African public -- A social media study
http://arxiv.org/abs/2006.06336
AUTHORS:        Vukosi Marivate ; Avashlin Moodley ; Athandiwe Saba
COMMENTS:       Under review for EMNLP 2020
HIGHLIGHT:      In this work, we expand on traditional media analysis by using Social Media discussions driven by or directed to South African government officials.

26, TITLE:      Spectral Image Segmentation with Global Appearance Modeling
http://arxiv.org/abs/2006.06573
AUTHORS:        Jeova F. S. Rocha Neto ; Pedro F. Felzenszwalb
HIGHLIGHT:      We introduce a new spectral method for image segmentation that incorporates long range relationships for global appearance modeling.

27, TITLE:      Mental Workload and Language Production in Non-Native Speaker IPA Interaction
http://arxiv.org/abs/2006.06331
AUTHORS:        Yunhan Wu ; Justin Edwards ; Orla Cooney ; Anna Bleakley ; Philip R. Doyle ; Leigh Clark ; Daniel Rough ; Benjamin R. Cowan
COMMENTS:       Accepted at CUI 2020
HIGHLIGHT:      We present a mixed-design experiment, wherein native (L1) and non-native (L2) English speakers completed tasks with IPAs through smartphones and smart speakers.

28, TITLE:      See what I'm saying? Comparing Intelligent Personal Assistant use for Native and Non-Native Language Speakers
http://arxiv.org/abs/2006.06328
AUTHORS:        Yunhan Wu ; Daniel Rough ; Anna Bleakley ; Justin Edwards ; Orla Cooney ; Philip R. Doyle ; Leigh Clark ; Benjamin R. Cowan
COMMENTS:       Accepted to Mobile HCI 2020
HIGHLIGHT:      Through native (L1) and non-native (L2) English speakers interacting with Google Assistant on a smartphone and smart speaker, we aim to understand this more deeply.

29, TITLE:      Learning a Unified Sample Weighting Network for Object Detection
http://arxiv.org/abs/2006.06568
AUTHORS:        Qi Cai ; Yingwei Pan ; Yu Wang ; Jingen Liu ; Ting Yao ; Tao Mei
COMMENTS:       CVPR 2020; The source code and model are publicly available at: \url{https://github.com/caiqi/sample-weighting-network}
HIGHLIGHT:      To this end, we devise a general loss function to cover most region-based object detectors with various sampling strategies, and then based on it we propose a unified sample weighting network to predict a sample's task weights.

30, TITLE:      Exploring Category-Agnostic Clusters for Open-Set Domain Adaptation
http://arxiv.org/abs/2006.06567
AUTHORS:        Yingwei Pan ; Ting Yao ; Yehao Li ; Chong-Wah Ngo ; Tao Mei
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we address this problem by augmenting the state-of-the-art domain adaptation technique, Self-Ensembling, with category-agnostic clusters in target domain.

31, TITLE:      CoMIR: Contrastive Multimodal Image Representation for Registration
http://arxiv.org/abs/2006.06325
AUTHORS:        Nicolas Pielawski ; Elisabeth Wetzer ; Johan Öfverstedt ; Jiahao Lu ; Carolina Wählby ; Joakim Lindblad ; Nataša Sladoje
COMMENTS:       21 pages, 11 figures
HIGHLIGHT:      We propose contrastive coding to learn shared, dense image representations, referred to as CoMIRs (Contrastive Multimodal Image Representations).

32, TITLE:      Surveys without Questions: A Reinforcement Learning Approach
http://arxiv.org/abs/2006.06323
AUTHORS:        Atanu R Sinha ; Deepali Jain ; Nikhil Sheoran ; Sopan Khosla ; Reshmi Sasidharan
COMMENTS:       The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19)
HIGHLIGHT:      We introduce a new way to interpret values generated by the value function of RL, as proxy ratings.

33, TITLE:      Hypernetwork-Based Augmentation
http://arxiv.org/abs/2006.06320
AUTHORS:        Chih-Yang Chen ; Che-Han Chang ; Edward Y. Chang
HIGHLIGHT:      In this paper, we propose an efficient gradient-based search algorithm, called Hypernetwork-Based Augmentation (HBA), which simultaneously learns model parameters and augmentation hyperparameters in a single training.

34, TITLE:      A Deep Learning Framework for Recognizing both Static and Dynamic Gestures
http://arxiv.org/abs/2006.06321
AUTHORS:        Osama Mazhar ; Sofiane Ramdani ; Andrea Cherubini
COMMENTS:       9 pages
HIGHLIGHT:      In this paper, we propose a unified framework that recognizes both static and dynamic gestures, using simple RGB vision (without depth sensing).

35, TITLE:      A Primer on Large Intelligent Surface (LIS) for Wireless Sensing in an Industrial Setting
http://arxiv.org/abs/2006.06563
AUTHORS:        Cristian J. Vaca-Rubio ; Pablo Ramirez-Espinosa ; Robin Jess Williams ; Kimmo Kansanen ; Zheng-Hua Tan ; Elisabeth de Carvalho ; Petar Popovski
HIGHLIGHT:      This paper addresses the potential of communication-sensing integration of Large Intelligent Surfaces (LIS) in an exemplary Industry 4.0 scenario.

36, TITLE:      RTEX: A novel methodology for Ranking, Tagging, and Explanatory diagnostic captioning of radiography exams
http://arxiv.org/abs/2006.06316
AUTHORS:        Vasiliki Kougia ; John Pavlopoulos ; Panagiotis Papapetrou ; Max Gordon
HIGHLIGHT:      This paper introduces RTEx, a novel methodology for a) ranking radiography exams based on their probability to contain an abnormality, b) generating abnormality tags for abnormal exams, and c) providing a diagnostic explanation in natural language for each abnormal exam.

37, TITLE:      Avoiding Side Effects in Complex Environments
http://arxiv.org/abs/2006.06547
AUTHORS:        Alexander Matt Turner ; Neale Ratzlaff ; Prasad Tadepalli
COMMENTS:       16 pages with appendices
HIGHLIGHT:      Avoiding Side Effects in Complex Environments

38, TITLE:      Attentive WaveBlock: Complementarity-enhanced Mutual Networks for Unsupervised Domain Adaptation in Person Re-identification
http://arxiv.org/abs/2006.06525
AUTHORS:        Wenhao Wang ; Fang Zhao ; Shengcai Liao ; Ling Shao
HIGHLIGHT:      In this paper, we propose a novel light-weight module, the Attentive WaveBlock (AWB), which can be integrated into the dual networks of mutual learning to enhance the complementarity and further depress noise in the pseudo-labels.

39, TITLE:      Rethinking the Truly Unsupervised Image-to-Image Translation
http://arxiv.org/abs/2006.06500
AUTHORS:        Kyungjune Baek ; Yunjey Choi ; Youngjung Uh ; Jaejun Yoo ; Hyunjung Shim
HIGHLIGHT:      In this paper, we tackle image-to-image translation in a fully unsupervised setting, i.e., neither paired images nor domain labels.

40, TITLE:      Transparency in Language Generation: Levels of Automation
http://arxiv.org/abs/2006.06295
AUTHORS:        Justin Edwards ; Allison Perrone ; Philip R. Doyle
COMMENTS:       Accepted for publication at CUI 2020
HIGHLIGHT:      We propose a taxonomy of language automation, based on the SAE levels of driving automation, to establish a shared set of terms for describing automated language.

41, TITLE:      Joint Training of Variational Auto-Encoder and Latent Energy-Based Model
http://arxiv.org/abs/2006.06059
AUTHORS:        Tian Han ; Erik Nijkamp ; Linqi Zhou ; Bo Pang ; Song-Chun Zhu ; Ying Nian Wu
HIGHLIGHT:      This paper proposes a joint training method to learn both the variational auto-encoder (VAE) and the latent energy-based model (EBM).

42, TITLE:      Scalable Partial Explainability in Neural Networks via Flexible Activation Functions
http://arxiv.org/abs/2006.06057
AUTHORS:        Schyler C. Sun ; Chen Li ; Zhuangkun Wei ; Antonios Tsourdos ; Weisi Guo
HIGHLIGHT:      In this paper, we achieve partially explainable learning model by symbolically explaining the role of activation functions (AF) under a scalable topology.

43, TITLE:      Marginal Utility for Planning in Continuous or Large Discrete Action Spaces
http://arxiv.org/abs/2006.06054
AUTHORS:        Zaheen Farraz Ahmad ; Levi H. S. Lelis ; Michael Bowling
HIGHLIGHT:      In this paper we explore explicitly learning a candidate action generator by optimizing a novel objective, marginal utility.

44, TITLE:      Fast Coherent Point Drift
http://arxiv.org/abs/2006.06281
AUTHORS:        Xiang-Wei Feng ; Da-Zheng Feng ; Yun Zhu
HIGHLIGHT:      By introducing a simple corresponding constraint, we develop a fast implementation of CPD.

45, TITLE:      A Novel Meta-Heuristic Optimization Algorithm Inspired by the Spread of Viruses
http://arxiv.org/abs/2006.06282
AUTHORS:        Zhixi Li ; Vincent Tam
HIGHLIGHT:      In this paper, a novel nature-inspired meta-heuristic optimization algorithm called virus spread optimization (VSO) is proposed.

46, TITLE:      Map3D: Registration Based Multi-Object Tracking on 3D Serial Whole Slide Images
http://arxiv.org/abs/2006.06038
AUTHORS:        Ruining Deng ; Haichun Yang ; Aadarsh Jha ; Yuzhe Lu ; Peng Chu ; Agnes Fogo ; Yuankai Huo
HIGHLIGHT:      In this paper, we propose a novel Multi-Object Association for Pathology in 3D (Map3D) method for automatically identifying and associating large-scale cross-sections of 3D objects from routine serial sectioning and WSI.

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

48, TITLE:      W-net: Simultaneous segmentation of multi-anatomical retinal structures using a multi-task deep neural network
http://arxiv.org/abs/2006.06277
AUTHORS:        Hongwei Zhao ; Chengtao Peng ; Lei Liu ; Bin Li
HIGHLIGHT:      In this study, we proposed a $\mathcal{W}$-net to simultaneously segment both the optic disc (OD) and the exudates in retinal images based on the multi-task learning (MTL) scheme.

49, TITLE:      XiaoiceSing: A High-Quality and Integrated Singing Voice Synthesis System
http://arxiv.org/abs/2006.06261
AUTHORS:        Peiling Lu ; Jie Wu ; Jian Luan ; Xu Tan ; Li Zhou
HIGHLIGHT:      This paper presents XiaoiceSing, a high-quality singing voice synthesis system which employs an integrated network for spectrum, F0 and duration modeling.

50, TITLE:      Towards Robust Fine-grained Recognition by Maximal Separation of Discriminative Features
http://arxiv.org/abs/2006.06028
AUTHORS:        Krishna Kanth Nakka ; Mathieu Salzmann
HIGHLIGHT:      In this paper, we identify the proximity of the latent representations of different classes in fine-grained recognition networks as a key factor to the success of adversarial attacks.

51, TITLE:      Report from the NSF Future Directions Workshop, Toward User-Oriented Agents: Research Directions and Challenges
http://arxiv.org/abs/2006.06026
AUTHORS:        Maxine Eskenazi ; Tiancheng Zhao
COMMENTS:       Final report of the NSF Future Directions Workshop, Toward User-Oriented Agents: Research Directions and Challenges
HIGHLIGHT:      Any opinions, findings and conclusions or future directions expressed in this document are those of the authors and do not necessarily reflect the views of the National Science Foundation.

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

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

54, TITLE:      Protecting Against Image Translation Deepfakes by Leaking Universal Perturbations from Black-Box Neural Networks
http://arxiv.org/abs/2006.06493
AUTHORS:        Nataniel Ruiz ; Sarah Adel Bargal ; Stan Sclaroff
HIGHLIGHT:      In this work, we develop efficient disruptions of black-box image translation deepfake generation systems.

55, TITLE:      Augmenting Data for Sarcasm Detection with Unlabeled Conversation Context
http://arxiv.org/abs/2006.06259
AUTHORS:        Hankyol Lee ; Youngjae Yu ; Gunhee Kim
HIGHLIGHT:      We present a novel data augmentation technique, CRA (Contextual Response Augmentation), which utilizes conversational context to generate meaningful samples for training.

56, TITLE:      Revisiting visual-inertial structure from motion for odometry and SLAM initialization
http://arxiv.org/abs/2006.06017
AUTHORS:        Georgios Evangelidis ; Branislav Micusik
HIGHLIGHT:      In this paper, an efficient closed-form solution for the state initialization in visual-inertial odometry (VIO) and simultaneous localization and mapping (SLAM) is presented.

57, TITLE:      Stochastic Segmentation Networks: Modelling Spatially Correlated Aleatoric Uncertainty
http://arxiv.org/abs/2006.06015
AUTHORS:        Miguel Monteiro ; Loïc Le Folgoc ; Daniel Coelho de Castro ; Nick Pawlowski ; Bernardo Marques ; Konstantinos Kamnitsas ; Mark van der Wilk ; Ben Glocker
COMMENTS:       17 pages, 11 figures, 2 tables
HIGHLIGHT:      In this paper, we introduce stochastic segmentation networks (SSNs), an efficient probabilistic method for modelling aleatoric uncertainty with any image segmentation network architecture.

58, TITLE:      Blissful Ignorance: Anti-Transfer Learning for Task Invariance
http://arxiv.org/abs/2006.06494
AUTHORS:        Eric Guizzo ; Tillman Weyde ; Giacomo Tarroni
COMMENTS:       NeurIPS2020
HIGHLIGHT:      We introduce the novel concept of anti-transfer learning for neural networks.

59, TITLE:      Graph Neural Networks for Motion Planning
http://arxiv.org/abs/2006.06248
AUTHORS:        Arbaaz Khan ; Alejandro Ribeiro ; Vijay Kumar ; Anthony G. Francis
HIGHLIGHT:      We present two techniques, GNNs over dense fixed graphs for low-dimensional problems and sampling-based GNNs for high-dimensional problems.

60, TITLE:      Privacy-Aware Activity Classification from First Person Office Videos
http://arxiv.org/abs/2006.06246
AUTHORS:        Partho Ghosh ; Md. Abrar Istiak ; Nayeeb Rashid ; Ahsan Habib Akash ; Ridwan Abrar ; Ankan Ghosh Dastider ; Asif Shahriyar Sushmit ; Taufiq Hasan
HIGHLIGHT:      In this work, we developed a privacy-aware activity classification system focusing on office videos.

61, TITLE:      CLEval: Character-Level Evaluation for Text Detection and Recognition Tasks
http://arxiv.org/abs/2006.06244
AUTHORS:        Youngmin Baek ; Daehyun Nam ; Sungrae Park ; Junyeop Lee ; Seung Shin ; Jeonghun Baek ; Chae Young Lee ; Hwalsuk Lee
COMMENTS:       12 pages, 8 figures
HIGHLIGHT:      Based on the fact that character is a key element of text, we hereby propose a Character-Level Evaluation metric (CLEval).

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

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

64, TITLE:      Walsh functions, scrambled $(0,m,s)$-nets, and negative covariance: applying symbolic computation to quasi-Monte Carlo integration
http://arxiv.org/abs/2006.06225
AUTHORS:        Jaspar Wiart ; Elaine Wong
COMMENTS:       27 pages; Supplementary material at https://wongey.github.io/digital-nets-walsh/
HIGHLIGHT:      We investigate base $b$ Walsh functions for which the variance of the integral estimator based on a scrambled $(0,m,s)$-net in base $b$ is less than or equal to that of the Monte-Carlo estimator based on the same number of points.

65, TITLE:      Discrete Latent Variable Representations for Low-Resource Text Classification
http://arxiv.org/abs/2006.06226
AUTHORS:        Shuning Jin ; Sam Wiseman ; Karl Stratos ; Karen Livescu
COMMENTS:       ACL 2020
HIGHLIGHT:      We consider several approaches to learning discrete latent variable models for text in the case where exact marginalization over these variables is intractable.

66, TITLE:      Deep Differential System Stability -- Learning advanced computations from examples
http://arxiv.org/abs/2006.06462
AUTHORS:        François Charton ; Amaury Hayat ; Guillaume Lample
HIGHLIGHT:      Using transformers over large generated datasets, we train models to learn properties of differential systems, such as local stability, behavior at infinity and controllability.

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

68, TITLE:      SECure: A Social and Environmental Certificate for AI Systems
http://arxiv.org/abs/2006.06217
AUTHORS:        Abhishek Gupta ; Camylle Lanteigne ; Sara Kingsley
COMMENTS:       Accepted for presentation at the Canadian Society for Ecological Economics 2020 Research Symposium
HIGHLIGHT:      This work proposes an ESG-inspired framework combining socio-technical measures to build eco-socially responsible AI systems.

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

70, TITLE:      System to Integrate Fairness Transparently: An Industry Approach
http://arxiv.org/abs/2006.06082
AUTHORS:        Emily Dodwell ; Cheryl Flynn ; Balachander Krishnamurthy ; Subhabrata Majumdar ; Ritwik Mitra
COMMENTS:       11 pages, 2 figures
HIGHLIGHT:      We propose a framework for industrial uses that addresses their methodological and mechanization needs.

71, TITLE:      DivNoising: Diversity Denoising with Fully Convolutional Variational Autoencoders
http://arxiv.org/abs/2006.06072
AUTHORS:        Mangal Prakash ; Alexander Krull ; Florian Jug
COMMENTS:       11 pages, 13 pages supplement, 4 figures, 19 supplementary figures
HIGHLIGHT:      Here, we propose DivNoising -- a denoising approach based on fully-convolutional variational autoencoders, overcoming this problem by predicting a whole distribution of denoised images.

72, TITLE:      S-semantics -- an example
http://arxiv.org/abs/2006.06077
AUTHORS:        Włodzimierz Drabent
COMMENTS:       13 pages, 1 figure
HIGHLIGHT:      Here we apply s-semantics to prove correctness and completeness of Fr\"uhwirth's $n$ queens program.

73, TITLE:      Deterministic Gaussian Averaged Neural Networks
http://arxiv.org/abs/2006.06061
AUTHORS:        Ryan Campbell ; Chris Finlay ; Adam M Oberman
HIGHLIGHT:      We present a deterministic method to compute the Gaussian average of neural networks used in regression and classification.

74, TITLE:      Improving Deep Metric Learning with Virtual Classes and Examples Mining
http://arxiv.org/abs/2006.06611
AUTHORS:        Pierre Jacob ; David Picard ; Aymeric Histace ; Edouard Klein
HIGHLIGHT:      To tackle this issue, we introduce MIRAGE, a generation-based method that relies on virtual classes entirely composed of generated examples that act as buffer areas between the training classes.

75, TITLE:      What makes instance discrimination good for transfer learning?
http://arxiv.org/abs/2006.06606
AUTHORS:        Nanxuan Zhao ; Zhirong Wu ; Rynson W. H. Lau ; Stephen Lin
HIGHLIGHT:      In this work, we investigate the following problems: What makes instance discrimination pretraining good for transfer learning?

76, TITLE:      Teaching Pre-Trained Models to Systematically Reason Over Implicit Knowledge
http://arxiv.org/abs/2006.06609
AUTHORS:        Alon Talmor ; Oyvind Tafjord ; Peter Clark ; Yoav Goldberg ; Jonathan Berant
HIGHLIGHT:      In this work, we provide a first demonstration that LMs can be trained to reliably perform systematic reasoning combining both implicit, pre-trained knowledge and explicit natural language statements.

77, TITLE:      Zeroth-Order Supervised Policy Improvement
http://arxiv.org/abs/2006.06600
AUTHORS:        Hao Sun ; Ziping Xu ; Yuhang Song ; Meng Fang ; Jiechao Xiong ; Bo Dai ; Zhengyou Zhang ; Bolei Zhou
HIGHLIGHT:      In this work, we propose a method referred to as Zeroth-Order Supervised Policy Improvement (ZOSPI) that exploits the estimated value function Q globally while preserves the local exploitation of the policy gradient methods.

78, TITLE:      Ensuring smoothly navigable approximation sets by Bezier curve parameterizations in evolutionary bi-objective optimization -- applied to brachytherapy treatment planning for prostate cancer
http://arxiv.org/abs/2006.06449
AUTHORS:        S. C. Maree ; T. Alderliesten ; P. A. N. Bosman
COMMENTS:       PPSN2020 (Parallel Problem Solving from Nature)
HIGHLIGHT:      In this work, we aim to improve approximation set navigability by enforcing a form of smoothness or continuity between solutions in terms of their decision variables.

79, TITLE:      Convolutional neural networks compression with low rank and sparse tensor decompositions
http://arxiv.org/abs/2006.06443
AUTHORS:        Pavel Kaloshin
HIGHLIGHT:      In this work, we consider a neural network compression method based on tensor decompositions.

80, TITLE:      Fall Detector Adapted to Nursing Home Needs through an Optical-Flow based CNN
http://arxiv.org/abs/2006.06201
AUTHORS:        Alexy Carlier ; Paul Peyramaure ; Ketty Favre ; Muriel Pressigout
HIGHLIGHT:      This work presents the requirements from the medical side and how it impacts the tuning of a CNN.

81, TITLE:      A Monolingual Approach to Contextualized Word Embeddings for Mid-Resource Languages
http://arxiv.org/abs/2006.06202
AUTHORS:        Pedro Ortiz Suárez ; Laurent Romary ; Benoît Sagot
HIGHLIGHT:      We use the multilingual OSCAR corpus, extracted from Common Crawl via language classification, filtering and cleaning, to train monolingual contextualized word embeddings (ELMo) for several mid-resource languages.

82, TITLE:      Unsupervised Learning of 3D Point Set Registration
http://arxiv.org/abs/2006.06200
AUTHORS:        Lingjing Wang ; Xiang Li ; Yi Fang
HIGHLIGHT:      Recent works leverage the power of deep learning for registering a pair of point sets.

83, TITLE:      GAIT-prop: A biologically plausible learning rule derived from backpropagation of error
http://arxiv.org/abs/2006.06438
AUTHORS:        Nasir Ahmad ; Marcel A. J. van Gerven ; Luca Ambrogioni
COMMENTS:       12 pages, 4 figures
HIGHLIGHT:      Traditional backpropagation of error, though a highly successful algorithm for learning in artificial neural network models, includes features which are biologically implausible for learning in real neural circuits.

84, TITLE:      A Probabilistic Model with Commonsense Constraints for Pattern-based Temporal Fact Extraction
http://arxiv.org/abs/2006.06436
AUTHORS:        Yang Zhou ; Tong Zhao ; Meng Jiang
COMMENTS:       7 pages, 1 figure
HIGHLIGHT:      In this work, we propose a probabilistic graphical model which formulates fact extraction in a generative process.

85, TITLE:      Complementary Visual Neuronal Systems Model for Collision Sensing
http://arxiv.org/abs/2006.06431
AUTHORS:        Qinbing Fu ; Shigang Yue
COMMENTS:       7 pages, 6 figures. This work has been accepted for publication in a future IEEE conference. Copyright has been transferred to the IEEE. This version may no longer be accessible after the conference publication in IEEE Xplore
HIGHLIGHT:      To fill this vacancy, we introduce a hybrid model combining two LGMDs (LGMD-1 and LGMD-2) with horizontally (rightward and leftward) sensitive LPTCs (LPTC-R and LPTC-L) specialising in fast collision perception.

86, TITLE:      Understanding Human Hands in Contact at Internet Scale
http://arxiv.org/abs/2006.06669
AUTHORS:        Dandan Shan ; Jiaqi Geng ; Michelle Shu ; David F. Fouhey
COMMENTS:       To appear at CVPR 2020 (Oral). Project and dataset webpage: http://fouheylab.eecs.umich.edu/~dandans/projects/100DOH/
HIGHLIGHT:      This paper proposes steps towards this by inferring a rich representation of hands engaged in interaction method that includes: hand location, side, contact state, and a box around the object in contact.

87, TITLE:      Disentangled Non-Local Neural Networks
http://arxiv.org/abs/2006.06668
AUTHORS:        Minghao Yin ; Zhuliang Yao ; Yue Cao ; Xiu Li ; Zheng Zhang ; Stephen Lin ; Han Hu
HIGHLIGHT:      Based on these findings, we present the disentangled non-local block, where the two terms are decoupled to facilitate learning for both terms.

88, TITLE:      VirTex: Learning Visual Representations from Textual Annotations
http://arxiv.org/abs/2006.06666
AUTHORS:        Karan Desai ; Justin Johnson
COMMENTS:       Code available at https://github.com/kdexd/virtex
HIGHLIGHT:      We propose VirTex -- a pretraining approach using semantically dense captions to learn visual representations.

89, TITLE:      Quasi-Dense Instance Similarity Learning
http://arxiv.org/abs/2006.06664
AUTHORS:        Jiangmiao Pang ; Linlu Qiu ; Haofeng Chen ; Qi Li ; Trevor Darrell ; Fisher Yu
COMMENTS:       SOTAs on multiple object tracking and one-shot object detection
HIGHLIGHT:      In this paper, we present a simple yet effective quasi-dense matching method to learn instance similarity from hundreds of region proposals in a pair of images.

90, TITLE:      Robust Multi-object Matching via Iterative Reweighting of the Graph Connection Laplacian
http://arxiv.org/abs/2006.06658
AUTHORS:        Yunpeng Shi ; Shaohan Li ; Gilad Lerman
HIGHLIGHT:      We propose an efficient and robust iterative solution to the multi-object matching problem.

91, TITLE:      Directional convergence and alignment in deep learning
http://arxiv.org/abs/2006.06657
AUTHORS:        Ziwei Ji ; Matus Telgarsky
HIGHLIGHT:      In this paper, we show that although the minimizers of cross-entropy and related classification losses are off at infinity, network weights learned by gradient flow converge in direction, with an immediate corollary that network predictions, training errors, and the margin distribution also converge.

92, TITLE:      Closed Loop Neural-Symbolic Learning via Integrating Neural Perception, Grammar Parsing, and Symbolic Reasoning
http://arxiv.org/abs/2006.06649
AUTHORS:        Qing Li ; Siyuan Huang ; Yining Hong ; Yixin Chen ; Ying Nian Wu ; Song-Chun Zhu
COMMENTS:       ICML 2020. Project page: https://liqing-ustc.github.io/NGS
HIGHLIGHT:      In this paper, we address these issues and close the loop of neural-symbolic learning by (1) introducing the \textbf{grammar} model as a \textit{symbolic prior} to bridge neural perception and symbolic reasoning, and (2) proposing a novel \textbf{back-search} algorithm which mimics the top-down human-like learning procedure to propagate the error through the symbolic reasoning module efficiently.

93, TITLE:      CoSDA-ML: Multi-Lingual Code-Switching Data Augmentation for Zero-Shot Cross-Lingual NLP
http://arxiv.org/abs/2006.06402
AUTHORS:        Libo Qin ; Minheng Ni ; Yue Zhang ; Wanxiang Che
COMMENTS:       Accepted at IJCAI2020. SOLE copyright holder is IJCAI (international Joint Conferences on Artificial Intelligence), all rights reserved. http://static.ijcai.org/2020-accepted_papers.html
HIGHLIGHT:      We propose a data augmentation framework to generate multi-lingual code-switching data to fine-tune mBERT, which encourages model to align representations from source and multiple target languages once by mixing their context information.

94, TITLE:      Interpretable Visualizations with Differentiating Embedding Networks
http://arxiv.org/abs/2006.06640
AUTHORS:        Isaac Robinson
COMMENTS:       10 pages, 4 figures, under review
HIGHLIGHT:      We present a visualization algorithm based on a novel unsupervised Siamese neural network training regime and loss function, called Differentiating Embedding Networks (DEN).

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

96, TITLE:      Privacy-Preserving Visual Feature Descriptors through Adversarial Affine Subspace Embedding
http://arxiv.org/abs/2006.06634
AUTHORS:        Mihai Dusmanu ; Johannes L. Schönberger ; Sudipta N. Sinha ; Marc Pollefeys
COMMENTS:       16 pages, 7 figures, 2 tables
HIGHLIGHT:      The core idea of our work is to drop constraints from each feature descriptor by embedding it within an affine subspace containing the original feature as well as one or more adversarial feature samples.

97, TITLE:      Petri Nets with Parameterised Data: Modelling and Verification (Extended Version)
http://arxiv.org/abs/2006.06630
AUTHORS:        Silvio Ghilardi ; Alessandro Gianola ; Marco Montali ; Andrey Rivkin
HIGHLIGHT:      In this work, we introduce and study an extension of coloured Petri nets, called catalog-nets, providing two key features to capture this type of processes.

98, TITLE:      Growing Artificial Neural Networks
http://arxiv.org/abs/2006.06629
AUTHORS:        John Mixter ; Ali Akoglu
COMMENTS:       14 pages, Accepted for publication in Springer Nature - Book Series: Transactions on Computational Science and Computational Intelligence, Advances in Artificial Intelligence and Applied Cognitive Computing - Springer ID: 89066307 (Book ID: 495585_1_En)
HIGHLIGHT:      We propose an algorithm, Artificial Neurogenesis (ANG), that grows rather than prunes the network and enables neural networks to be trained and executed in low SWaP embedded hardware.

99, TITLE:      Diagnosis and Analysis of Celiac Disease and Environmental Enteropathy on Biopsy Images using Deep Learning Approaches
http://arxiv.org/abs/2006.06627
AUTHORS:        Kamran Kowsari
COMMENTS:       PhD dissertation, Univ Virginia (May 2020)
HIGHLIGHT:      The dataset used in this research is collected from different centers with different staining standards.

100, TITLE:     Real-Time Video Inference on Edge Devices via Adaptive Model Streaming
http://arxiv.org/abs/2006.06628
AUTHORS:        Mehrdad Khani ; Pouya Hamadanian ; Arash Nasr-Esfahany ; Mohammad Alizadeh
HIGHLIGHT:      In this paper we propose Adaptive Model Streaming (AMS), a cloud-assisted approach to real-time video inference on edge devices.

101, TITLE:     Scalable Multi-Agent Reinforcement Learning for Networked Systems with Average Reward
http://arxiv.org/abs/2006.06626
AUTHORS:        Guannan Qu ; Yiheng Lin ; Adam Wierman ; Na Li
HIGHLIGHT:      In this paper, we identify a rich class of networked MARL problems where the model exhibits a local dependence structure that allows it to be solved in a scalable manner.

102, TITLE:     SLIC-UAV: A Method for monitoring recovery in tropical restoration projects through identification of signature species using UAVs
http://arxiv.org/abs/2006.06624
AUTHORS:        Jonathan Williams ; Carola-Bibiane Schönlieb ; Tom Swinfield ; Bambang Irawan ; Eva Achmad ; Muhammad Zudhi ;  Habibi ; Elva Gemita ; David A. Coomes
HIGHLIGHT:      Here we present a new pipeline, SLIC-UAV, for processing Unmanned Aerial Vehicle (UAV) imagery to map early-successional species in tropical forests.

103, TITLE:     From proprioception to long-horizon planning in novel environments: A hierarchical RL model
http://arxiv.org/abs/2006.06620
AUTHORS:        Nishad Gothoskar ; Miguel Lázaro-Gredilla ; Dileep George
HIGHLIGHT:      In this work, we introduce a simple, three-level hierarchical architecture that reflects these distinctions.

104, TITLE:     MatchGAN: A Self-Supervised Semi-Supervised Conditional Generative Adversarial Network
http://arxiv.org/abs/2006.06614
AUTHORS:        Jiaze Sun ; Binod Bhattarai ; Tae-Kyun Kim
HIGHLIGHT:      We propose a novel self-supervised semi-supervised learning approach for conditional Generative Adversarial Networks (GANs).



==========Updates to Previous Papers==========
1, TITLE:       Multi-task Batch Reinforcement Learning with Metric Learning
http://arxiv.org/abs/1909.11373
AUTHORS:        Jiachen Li ; Quan Vuong ; Shuang Liu ; Minghua Liu ; Kamil Ciosek ; Henrik Iskov Christensen ; Hao Su
HIGHLIGHT:      To robustify task inference, we propose a novel application of the triplet loss.

2, TITLE:       A Robust Method for Image Stitching
http://arxiv.org/abs/2004.03860
AUTHORS:        Matti Pellikka ; Valtteri Lahtinen
HIGHLIGHT:      We propose a novel method for large-scale image stitching that is robust against repetitive patterns and featureless regions in the imaginary.

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

4, TITLE:       Fact or Fiction: Verifying Scientific Claims
http://arxiv.org/abs/2004.14974
AUTHORS:        David Wadden ; Shanchuan Lin ; Kyle Lo ; Lucy Lu Wang ; Madeleine van Zuylen ; Arman Cohan ; Hannaneh Hajishirzi
COMMENTS:       16 pages (including appendices), 7 figures, 8 tables. GitHub: https://github.com/allenai/scifact
HIGHLIGHT:      We introduce scientific claim verification, a new task to select abstracts from the research literature containing evidence that supports or refutes a given scientific claim, and to identify rationales justifying each decision.

5, TITLE:       Single Image Deraining via Scale-space Invariant Attention Neural Network
http://arxiv.org/abs/2006.05049
AUTHORS:        Bo Pang ; Deming Zhai ; Junjun Jiang ; Xianming Liu
HIGHLIGHT:      In this paper, we tackle the notion of scale that deals with visual changes in appearance of rain steaks with respect to the camera.

6, TITLE:       Dynamic Fusion Network for Multi-Domain End-to-end Task-Oriented Dialog
http://arxiv.org/abs/2004.11019
AUTHORS:        Libo Qin ; Xiao Xu ; Wanxiang Che ; Yue Zhang ; Ting Liu
COMMENTS:       ACL2020
HIGHLIGHT:      To this end, we investigate methods that can make explicit use of domain knowledge and introduce a shared-private network to learn shared and specific knowledge.

7, TITLE:       Predictive Coding Approximates Backprop along Arbitrary Computation Graphs
http://arxiv.org/abs/2006.04182
AUTHORS:        Beren Millidge ; Alexander Tschantz ; Christopher L. Buckley
COMMENTS:       Submitted to NeurIPS 2020. Updated Acknowledgements. 11/06/020 -- fixed typos in maths
HIGHLIGHT:      Here, we demonstrate that predictive coding converges asymptotically (and in practice rapidly) to exact backprop gradients on arbitrary computation graphs using only local learning rules.

8, TITLE:       UHH-LT at SemEval-2020 Task 12: Fine-Tuning of Pre-Trained Transformer Networks for Offensive Language Detection
http://arxiv.org/abs/2004.11493
AUTHORS:        Gregor Wiedemann ; Seid Muhie Yimam ; Chris Biemann
HIGHLIGHT:      In this paper, we compare current pre-trained transformer networks with and without MLM fine-tuning on their performance for offensive language detection.

9, TITLE:       DSAC: Distributional Soft Actor Critic for Risk-Sensitive Reinforcement Learning
http://arxiv.org/abs/2004.14547
AUTHORS:        Xiaoteng Ma ; Li Xia ; Zhengyuan Zhou ; Jun Yang ; Qianchuan Zhao
HIGHLIGHT:      In this paper, we present a new reinforcement learning (RL) algorithm called Distributional Soft Actor Critic (DSAC), which exploits the distributional information of accumulated rewards to achieve better performance.

10, TITLE:      KRED: Knowledge-Aware Document Representation for News Recommendations
http://arxiv.org/abs/1910.11494
AUTHORS:        Danyang Liu ; Jianxun Lian ; Shiyin Wang ; Ying Qiao ; Jiun-Hung Chen ; Guangzhong Sun ; Xing Xie
HIGHLIGHT:      In this paper, we propose KRED, which is a fast and effective model to enhance arbitrary document representation with a knowledge graph.

11, TITLE:      Evolutionary Reinforcement Learning for Sample-Efficient Multiagent Coordination
http://arxiv.org/abs/1906.07315
AUTHORS:        Shauharda Khadka ; Somdeb Majumdar ; Santiago Miret ; Stephen McAleer ; Kagan Tumer
COMMENTS:       Proceedings of the 37th International Conference on Machine Learning, Vienna, Austria, PMLR 108, 2020
HIGHLIGHT:      We introduce Multiagent Evolutionary Reinforcement Learning (MERL), a split-level training platform that handles the two objectives separately through two optimization processes.

12, TITLE:      Fully Convolutional Online Tracking
http://arxiv.org/abs/2004.07109
AUTHORS:        Yutao Cui ; Cheng Jiang ; Limin Wang ; Gangshan Wu
COMMENTS:       FCOT achieves the best EAO of 0.508 on VOT2018. Code will be made available at https://github.com/MCG-NJU/FCOT
HIGHLIGHT:      To tackle this issue, we present the first fully convolutional online tracking framework (FCOT), with a focus on enabling online learning for both classification and regression branches.

13, TITLE:      RustHorn: CHC-based Verification for Rust Programs (full version)
http://arxiv.org/abs/2002.09002
AUTHORS:        Yusuke Matsushita ; Takeshi Tsukada ; Naoki Kobayashi
COMMENTS:       Full version of the same-titled paper in ESOP2020
HIGHLIGHT:      This paper proposes a novel translation of pointer-manipulating Rust programs into CHCs, which clears away pointers and memories by leveraging ownership.

14, TITLE:      Embed2Detect: Temporally Clustered Embedded Words for Event Detection in Social Media
http://arxiv.org/abs/2006.05908
AUTHORS:        Hansi Hettiarachchi ; Mariam Adedoyin-Olowe ; Jagdev Bhogal ; Mohamed Medhat Gaber
COMMENTS:       Submitted to Journal of Neural Computing and Applications, Springer
HIGHLIGHT:      In this paper, we propose a novel method termed Embed2Detect for event detection in social media by combining the characteristics in prediction-based word embeddings and hierarchical agglomerative clustering.

15, TITLE:      MemNet: Memory-Efficiency Guided Neural Architecture Search with Augment-Trim learning
http://arxiv.org/abs/1907.09569
AUTHORS:        Peiye Liu ; Bo Wu ; Huadong Ma ; Mingoo Seok
HIGHLIGHT:      To close this gap, we propose MemNet, an augment-trim learning-based neural network search framework that optimizes not only performance but also memory requirement.

16, TITLE:      Predicting Motion of Vulnerable Road Users using High-Definition Maps and Efficient ConvNets
http://arxiv.org/abs/1906.08469
AUTHORS:        Fang-Chieh Chou ; Tsung-Han Lin ; Henggang Cui ; Vladan Radosavljevic ; Thi Nguyen ; Tzu-Kuo Huang ; Matthew Niedoba ; Jeff Schneider ; Nemanja Djuric
COMMENTS:       Accepted for publication at IEEE Intelligent Vehicles Symposium (IV) 2020
HIGHLIGHT:      To address this issue, in the current study we present a deep learning-based method for predicting VRU movement, where we rasterize high-definition maps and actor's surroundings into a bird's-eye view image used as an input to deep convolutional networks.

17, TITLE:      Towards an Intrinsic Definition of Robustness for a Classifier
http://arxiv.org/abs/2006.05095
AUTHORS:        Théo Giraudon ; Vincent Gripon ; Matthias Löwe ; Franck Vermet
COMMENTS:       13 pages
HIGHLIGHT:      In this paper, we point out that averaging the radius of robustness of samples in a validation set is a statistically weak measure.

18, TITLE:      The Effect of Moderation on Online Mental Health Conversations
http://arxiv.org/abs/2005.09225
AUTHORS:        David Wadden ; Tal August ; Qisheng Li ; Tim Althoff
COMMENTS:       13 pages, 12 figures. 3 tables
HIGHLIGHT:      In this work, we leveraged a natural experiment, occurring across 200,000 messages from 7,000 conversations hosted on a mental health mobile application, to evaluate the effects of moderation on online mental health discussions.

19, TITLE:      Improving Generalized Zero-Shot Learning by Semantic Discriminator
http://arxiv.org/abs/2005.13956
AUTHORS:        Xinpeng Li
HIGHLIGHT:      We propose a new approach to distinguish whether the instances come from the seen or unseen classes.

20, TITLE:      Comparison of State-of-the-Art Deep Learning APIs for Image Multi-Label Classification using Semantic Metrics
http://arxiv.org/abs/1903.09190
AUTHORS:        Adam Kubany ; Shimon Ben Ishay ; Ruben-sacha Ohayon ; Armin Shmilovici ; Lior Rokach ; Tomer Doitshman
HIGHLIGHT:      In this study, we evaluate and compare the performance of 13 of the most prominent commercial and open-source APIs in a best-of-breed challenge on the Visual Genome and Open Images benchmark datasets.

21, TITLE:      Multiple target tracking based on sets of trajectories
http://arxiv.org/abs/1605.08163
AUTHORS:        Ángel F. García-Fernández ; Lennart Svensson ; Mark R. Morelande
COMMENTS:       MATLAB implementations of algorithms based on sets of trajectories can be found at https://github.com/Agarciafernandez
HIGHLIGHT:      We propose a solution of the multiple target tracking (MTT) problem based on sets of trajectories and the random finite set framework.

22, TITLE:      A Quantitative History of A.I. Research in the United States and China
http://arxiv.org/abs/2003.02763
AUTHORS:        Daniel Ish ; Andrew Lohn ; Christian Curriden
HIGHLIGHT:      Motivated by recent interest in the status and consequences of competition between the U.S. and China in A.I. research, we analyze 60 years of abstract data scraped from Scopus to explore and quantify trends in publications on A.I. topics from institutions affiliated with each country.

23, TITLE:      DenoiSeg: Joint Denoising and Segmentation
http://arxiv.org/abs/2005.02987
AUTHORS:        Tim-Oliver Buchholz ; Mangal Prakash ; Alexander Krull ; Florian Jug
COMMENTS:       10 pages, 4 figures, 2 pages supplement (4 figures)
HIGHLIGHT:      Here we propose DenoiSeg, a new method that can be trained end-to-end on only a few annotated ground truth segmentations.

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

25, TITLE:      VerSe: A Vertebrae Labelling and Segmentation Benchmark
http://arxiv.org/abs/2001.09193
AUTHORS:        Anjany Sekuboyina ; Amirhossein Bayat ; Malek E. Husseini ; Maximilian Löffler ; Markus Rempfler ; Jan Kukačka ; Giles Tetteh ; Alexander Valentinitsch ; Christian Payer ; Darko Štern ; Martin Urschler ; Maodong Chen ; Dalong Cheng ; Nikolas Lessmann ; Yujin Hu ; Tianfu Wang ; Dong Yang ; Daguang Xu ; Felix Ambellan ; Tamaz Amiranashvili ; Moritz Ehlke ; Hans Lamecker ; Sebastian Lehnert ; Marilia Lirio ; Nicolás Pérez de Olaguer ; Heiko Ramm ; Manish Sahu ; Alexander Tack ; Stefan Zachow ; Tao Jiang ; Xinjun Ma ; Christoph Angerman ; Xin Wang ; Qingyue Wei ; Kevin Brown ; Matthias Wolf ; Alexandre Kirszenberg ; Élodie Puybareauq ; Björn H. Menze ; Jan S. Kirschke
HIGHLIGHT:      This work presents a detailed performance analysis of these algorithms with the best performing algorithm achieving a vertebrae identification rate of 95% and a Dice coefficient of 90%.

26, TITLE:      Distinguishing noisy boson sampling from classical simulations
http://arxiv.org/abs/1905.11458
AUTHORS:        Valery Shchesnovich
COMMENTS:       18 pages (6 pages of the main text), two figures (both in colour)
HIGHLIGHT:      In this work it is shown that one can efficiently distinguish the output distribution of such a noisy boson sampling from any approximation accounting for the low-order quantum multiboson interferences, which includes the mentioned classical algorithms.

27, TITLE:      Machine Learning Automatically Detects COVID-19 using Chest CTs in a Large Multicenter Cohort
http://arxiv.org/abs/2006.04998
AUTHORS:        Bogdan Georgescu ; Shikha Chaganti ; Gorka Bastarrika Aleman ; Eduardo Jose Mortani Barbosa Jr. ; Jordi Broncano Cabrero ; Guillaume Chabin ; Thomas Flohr ; Philippe Grenier ; Sasa Grbic ; Nakul Gupta ; François Mellot ; Savvas Nicolaou ; Thomas Re ; Pina Sanelli ; Alexander W. Sauter ; Youngjin Yoo ; Valentin Ziebandt ; Dorin Comaniciu
HIGHLIGHT:      Machine Learning Automatically Detects COVID-19 using Chest CTs in a Large Multicenter Cohort

28, TITLE:      Stance Detection on Social Media: State of the Art and Trends
http://arxiv.org/abs/2006.03644
AUTHORS:        Abeer AlDayel ; Walid Magdy
HIGHLIGHT:      An exhaustive review of stance detection techniques on social media is presented,including the task definition, the different types of targets in stance detection, the features set used, and the various machine learning approaches applied.

29, TITLE:      Proceedings of the Artificial Intelligence for Cyber Security (AICS) Workshop 2020
http://arxiv.org/abs/2002.08320
AUTHORS:        Dennis Ross ; Arunesh Sinha ; Diane Staheli ; Bill Streilein
HIGHLIGHT:      The workshop will focus on the application of artificial intelligence to problems in cyber security.

30, TITLE:      Co-Heterogeneous and Adaptive Segmentation from Multi-Source and Multi-Phase CT Imaging Data: A Study on Pathological Liver and Lesion Segmentation
http://arxiv.org/abs/2005.13201
AUTHORS:        Ashwin Raju ; Chi-Tung Cheng ; Yunakai Huo ; Jinzheng Cai ; Junzhou Huang ; Jing Xiao ; Le Lu ; ChienHuang Liao ; Adam P Harrison
COMMENTS:       23 pages, 8 figures
HIGHLIGHT:      In this work, we present a novel segmentation strategy, co-heterogenous and adaptive segmentation (CHASe), which only requires a small labeled cohort of single phase imaging data to adapt to any unlabeled cohort of heterogenous multi-phase data with possibly new clinical scenarios and pathologies.

31, 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.

32, TITLE:      Sparse Interpolation With Errors in Chebyshev Basis Beyond Redundant-Block Decoding
http://arxiv.org/abs/1912.05719
AUTHORS:        Erich L. Kaltofen ; Zhi-Hong Yang
HIGHLIGHT:      We present sparse interpolation algorithms for recovering a polynomial with $\le B$ terms from $N$ evaluations at distinct values for the variable when $\le E$ of the evaluations can be erroneous.

33, TITLE:      Recurrent and Spiking Modeling of Sparse Surgical Kinematics
http://arxiv.org/abs/2005.05868
AUTHORS:        Neil Getty ; Zixuan Zhao ; Stephan Gruessner ; Liaohai Chen ; Fangfang Xia
COMMENTS:       5 pages, 8 figures, accepted ICONS 2020
HIGHLIGHT:      In this study, we explore the possibility of using only kinematic data to predict surgeons of similar skill levels.

34, TITLE:      Why X rather than Y? Explaining Neural Model' Predictions by Generating Intervention Counterfactual Samples
http://arxiv.org/abs/1911.02042
AUTHORS:        Thai Le ; Suhang Wang ; Dongwon Lee
HIGHLIGHT:      To mitigate this limitation, therefore, we borrow two notable ideas (i.e., "explanation by intervention" from causality and "explanation are contrastive" from philosophy) and propose a novel solution, named as GRACE, that better explains neural network models' predictions for tabular datasets.

35, TITLE:      Defending Against Universal Attacks Through Selective Feature Regeneration
http://arxiv.org/abs/1906.03444
AUTHORS:        Tejas Borkar ; Felix Heide ; Lina Karam
COMMENTS:       CVPR 2020. Code: https://github.com/tsborkar/Selective-feature-regeneration Webpage: https://www.cs.princeton.edu/~fheide/SelectiveFeatureRegeneration/
HIGHLIGHT:      Departing from existing defense strategies that work mostly in the image domain, we present a novel defense which operates in the DNN feature domain and effectively defends against such universal perturbations.

36, TITLE:      The Unstoppable Rise of Computational Linguistics in Deep Learning
http://arxiv.org/abs/2005.06420
AUTHORS:        James Henderson
COMMENTS:       13 pages. Accepted for publication at ACL 2020, in the theme track
HIGHLIGHT:      In this paper, we trace the history of neural networks applied to natural language understanding tasks, and identify key contributions which the nature of language has made to the development of neural network architectures.

37, TITLE:      ConfNet2Seq: Full Length Answer Generation from Spoken Questions
http://arxiv.org/abs/2006.05163
AUTHORS:        Vaishali Pal ; Manish Shrivastava ; Laurent Besacier
COMMENTS:       Accepted at Text, Speech and Dialogue, 2020
HIGHLIGHT:      We propose a novel system to generate full length natural language answers from spoken questions and factoid answers. We release a large-scale dataset of 259,788 samples of spoken questions, their factoid answers and corresponding full-length textual answers.

38, TITLE:      CompLex: A New Corpus for Lexical Complexity Prediction from Likert Scale Data
http://arxiv.org/abs/2003.07008
AUTHORS:        Matthew Shardlow ; Michael Cooper ; Marcos Zampieri
COMMENTS:       Proceedings of the 1st Workshop on Tools and Resources to Empower People with REAding DIfficulties (READI). pp. 57-62
HIGHLIGHT:      With a few exceptions, previous studies have approached the task as a binary classification task in which systems predict a complexity value (complex vs. non-complex) for a set of target words in a text.

39, TITLE:      A t-distribution based operator for enhancing out of distribution robustness of neural network classifiers
http://arxiv.org/abs/2006.05389
AUTHORS:        Niccolò Antonello ; Philip N. Garner
COMMENTS:       5 pages, 5 figures, to be published in IEEE Signal Processing Letters, reproducible code https://github.com/idiap/tsoftmax
HIGHLIGHT:      A t-distribution based operator for enhancing out of distribution robustness of neural network classifiers

40, TITLE:      Dynamic Refinement Network for Oriented and Densely Packed Object Detection
http://arxiv.org/abs/2005.09973
AUTHORS:        Xingjia Pan ; Yuqiang Ren ; Kekai Sheng ; Weiming Dong ; Haolei Yuan ; Xiaowei Guo ; Chongyang Ma ; Changsheng Xu
COMMENTS:       Accepted by CVPR 2020 as Oral
HIGHLIGHT:      To resolve the first two issues, we present a dynamic refinement network that consists of two novel components, i.e., a feature selection module (FSM) and a dynamic refinement head (DRH). To address the limited availability of related benchmarks, we collect an extensive and fully annotated dataset, namely, SKU110K-R, which is relabeled with oriented bounding boxes based on SKU110K.

41, TITLE:      SemEval-2019 Task 1: Cross-lingual Semantic Parsing with UCCA
http://arxiv.org/abs/1903.02953
AUTHORS:        Daniel Hershcovich ; Zohar Aizenbud ; Leshem Choshen ; Elior Sulem ; Ari Rappoport ; Omri Abend
COMMENTS:       SemEval 2019 Shared task. arXiv admin note: substantial text overlap with arXiv:1805.12386
HIGHLIGHT:      We present the SemEval 2019 shared task on UCCA parsing in English, German and French, and discuss the participating systems and results.

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

43, TITLE:      GANHopper: Multi-Hop GAN for Unsupervised Image-to-Image Translation
http://arxiv.org/abs/2002.10102
AUTHORS:        Wallace Lira ; Johannes Merz ; Daniel Ritchie ; Daniel Cohen-Or ; Hao Zhang
COMMENTS:       9 pages, 9 figures. Code is available at https://github.com/wallacemplira/ganhopper
HIGHLIGHT:      We introduce GANHopper, an unsupervised image-to-image translation network that transforms images gradually between two domains, through multiple hops.

44, TITLE:      An Abstraction-guided Approach to Scalable and Rigorous Floating-Point Error Analysis
http://arxiv.org/abs/2004.11960
AUTHORS:        Arnab Das ; Ian Briggs ; Ganesh Gopalakrishnan ; Sriram Krishnamoorthy
COMMENTS:       20 pages
HIGHLIGHT:      In this work, we present Satire, a new tool that sheds light on how scalability and bound-tightness can be attained through a combination of incremental analysis, abstraction, and judicious use of concrete and symbolic evaluation.

45, TITLE:      Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data
http://arxiv.org/abs/2006.01414
AUTHORS:        Xinyu Wang ; Yong Jiang ; Kewei Tu
COMMENTS:       IWPT 2020 shared task. After fixing the bug, our proposed parser performs better than the team that ranked 1st in the official results
HIGHLIGHT:      This paper presents the system used in our submission to the \textit{IWPT 2020 Shared Task}.

46, TITLE:      HybridPose: 6D Object Pose Estimation under Hybrid Representations
http://arxiv.org/abs/2001.01869
AUTHORS:        Chen Song ; Jiaru Song ; Qixing Huang
HIGHLIGHT:      We introduce HybridPose, a novel 6D object pose estimation approach.

47, TITLE:      Adversarial Item Promotion: Vulnerabilities at the Core of Top-N Recommenders that Use Images to Address Cold Start
http://arxiv.org/abs/2006.01888
AUTHORS:        Zhuoran Liu ; Martha Larson
COMMENTS:       Our code is available at https://github.com/liuzrcc/AIP
HIGHLIGHT:      In this paper, we demonstrate how unscrupulous merchants can create item images that artificially promote their products, improving their rankings.

48, TITLE:      Disease Detection from Lung X-ray Images based on Hybrid Deep Learning
http://arxiv.org/abs/2003.00682
AUTHORS:        Subrato Bharati ; Prajoy Podder ; M. Rubaiyat Hossain Mondal
COMMENTS:       20 figures
HIGHLIGHT:      In this paper, a chest X ray image dataset has been used in order to diagnosis properly and analysis the lung disease.

49, TITLE:      MultiXNet: Multiclass Multistage Multimodal Motion Prediction
http://arxiv.org/abs/2006.02000
AUTHORS:        Nemanja Djuric ; Henggang Cui ; Zhaoen Su ; Shangxuan Wu ; Huahua Wang ; Fang-Chieh Chou ; Luisa San Martin ; Song Feng ; Rui Hu ; Yang Xu ; Alyssa Dayan ; Sidney Zhang ; Brian C. Becker ; Gregory P. Meyer ; Carlos Vallespi-Gonzalez ; Carl K. Wellington
HIGHLIGHT:      To address this task we propose MultiXNet, an end-to-end approach for detection and motion prediction based directly on lidar sensor data.

50, TITLE:      ClarQ: A large-scale and diverse dataset for Clarification Question Generation
http://arxiv.org/abs/2006.05986
AUTHORS:        Vaibhav Kumar ; Alan W. black
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      In order to overcome these limitations, we devise a novel bootstrapping framework (based on self-supervision) that assists in the creation of a diverse, large-scale dataset of clarification questions based on post-comment tuples extracted from stackexchange. We release this dataset in order to foster research into the field of clarification question generation with the larger goal of enhancing dialog and question answering systems.

51, TITLE:      Deeply Shape-guided Instance Segmentation
http://arxiv.org/abs/1911.11263
AUTHORS:        Hao Ding ; Siyuan Qiao ; Alan Yuille ; Wei Shen
HIGHLIGHT:      To address this issue, we propose a Deeply Shape-guided (DSG) framework for instance segmentation.

52, TITLE:      AutoDNNchip: An Automated DNN Chip Predictor and Builder for Both FPGAs and ASICs
http://arxiv.org/abs/2001.03535
AUTHORS:        Pengfei Xu ; Xiaofan Zhang ; Cong Hao ; Yang Zhao ; Yongan Zhang ; Yue Wang ; Chaojian Li ; Zetong Guan ; Deming Chen ; Yingyan Lin
COMMENTS:       Accepted by 28th ACM/SIGDA International Symposium on Field-Programmable Gate Arrays (FPGA'2020)
HIGHLIGHT:      To enable fast and effective DNN chip design, we propose AutoDNNchip - a DNN chip generator that can automatically generate both FPGA- and ASIC-based DNN chip implementation given DNNs from machine learning frameworks (e.g., PyTorch) for a designated application and dataset.

53, TITLE:      Empirical Analysis of Zipf's Law, Power Law, and Lognormal Distributions in Medical Discharge Reports
http://arxiv.org/abs/2003.13352
AUTHORS:        Juan C Quiroz ; Liliana Laranjo ; Catalin Tufanaru ; Ahmet Baki Kocaballi ; Dana Rezazadegan ; Shlomo Berkovsky ; Enrico Coiera
COMMENTS:       Added a limitations paragraph and expanded discussion of Bayesian non-parametric implications
HIGHLIGHT:      We examined 20,000 medical discharge reports from the MIMIC-III dataset.

54, TITLE:      Metric-Learning-Assisted Domain Adaptation
http://arxiv.org/abs/2004.10963
AUTHORS:        Yueming Yin ; Zhen Yang ; Haifeng Hu ; Xiaofu Wu
HIGHLIGHT:      We thus propose a novel metric-learning-assisted domain adaptation (MLA-DA) method, which employs a novel triplet loss for helping better feature alignment.

55, TITLE:      Efficient Querying from Weighted Binary Codes
http://arxiv.org/abs/1912.05006
AUTHORS:        Zhenyu Weng ; Yuesheng Zhu
COMMENTS:       13 pages, accepted by AAAI2020
HIGHLIGHT:      In this paper, we propose a new method to rank the weighted binary codes and return the nearest weighted binary codes of the query efficiently.

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

57, TITLE:      Improved Few-Shot Visual Classification
http://arxiv.org/abs/1912.03432
AUTHORS:        Peyman Bateni ; Raghav Goyal ; Vaden Masrani ; Frank Wood ; Leonid Sigal
HIGHLIGHT:      In this paper, we explore the hypothesis that a simple class-covariance-based distance metric, namely the Mahalanobis distance, adopted into a state of the art few-shot learning approach (CNAPS) can, in and of itself, lead to a significant performance improvement.

58, TITLE:      SpotNet: Self-Attention Multi-Task Network for Object Detection
http://arxiv.org/abs/2002.05540
AUTHORS:        Hughes Perreault ; Guillaume-Alexandre Bilodeau ; Nicolas Saunier ; Maguelonne Héritier
HIGHLIGHT:      Using these labels, we train an object detection model to produce foreground/background segmentation maps as well as bounding boxes while sharing most model parameters.

59, TITLE:      DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing
http://arxiv.org/abs/1911.13225
AUTHORS:        Shaohui Liu ; Yinda Zhang ; Songyou Peng ; Boxin Shi ; Marc Pollefeys ; Zhaopeng Cui
COMMENTS:       Camera-ready version to appear in CVPR 2020. Project page: http://b1ueber2y.me/projects/DIST-Renderer
HIGHLIGHT:      We propose a differentiable sphere tracing algorithm to bridge the gap between inverse graphics methods and the recently proposed deep learning based implicit signed distance function.

60, TITLE:      Unsupervised Intra-domain Adaptation for Semantic Segmentation through Self-Supervision
http://arxiv.org/abs/2004.07703
AUTHORS:        Fei Pan ; Inkyu Shin ; Francois Rameau ; Seokju Lee ; In So Kweon
COMMENTS:       Accepted to CVPR 2020 as an Oral Presentation. Code is available at https://github.com/feipan664/IntraDA
HIGHLIGHT:      In this work, we propose a two-step self-supervised domain adaptation approach to minimize the inter-domain and intra-domain gap together.

61, TITLE:      Self-Supervised Reinforcement Learning for Recommender Systems
http://arxiv.org/abs/2006.05779
AUTHORS:        Xin Xin ; Alexandros Karatzoglou ; Ioannis Arapakis ; Joemon M. Jose
COMMENTS:       SIGIR2020
HIGHLIGHT:      In this paper, we propose self-supervised reinforcement learning for sequential recommendation tasks.

62, TITLE:      Chameleon: Learning Model Initializations Across Tasks With Different Schemas
http://arxiv.org/abs/1909.13576
AUTHORS:        Lukas Brinkmeyer ; Rafael Rego Drumond ; Randolf Scholz ; Josif Grabocka ; Lars Schmidt-Thieme
COMMENTS:       18 pages, 7 figures
HIGHLIGHT:      In this paper, we address the problem of meta-learning parameter initialization across tasks with different schemas, i.e., if the number of predictors varies across tasks, while they still share some variables.