2020.07.24.txt

==========New Papers==========
1, TITLE:       PP-YOLO: An Effective and Efficient Implementation of Object Detector
http://arxiv.org/abs/2007.12099
AUTHORS:        Xiang Long ; Kaipeng Deng ; Guanzhong Wang ; Yang Zhang ; Qingqing Dang ; Yuan Gao ; Hui Shen ; Jianguo Ren ; Shumin Han ; Errui Ding ; Shilei Wen
HIGHLIGHT:      The goal of this paper is to implement an object detector with relatively balanced effectiveness and efficiency that can be directly applied in actual application scenarios, rather than propose a novel detection model.

2, TITLE:       A Study on Evaluation Standard for Automatic Crack Detection Regard the Random Fractal
http://arxiv.org/abs/2007.12082
AUTHORS:        Hongyu Li ; Jihe Wang ; Yu Zhang ; Zirui Wang ; Tiejun Wang
HIGHLIGHT:      This paper presents a study on the evaluation standard.

3, TITLE:       NITS-Hinglish-SentiMix at SemEval-2020 Task 9: Sentiment Analysis For Code-Mixed Social Media Text
http://arxiv.org/abs/2007.12081
AUTHORS:        Subhra Jyoti Baroi ; Nivedita Singh ; Ringki Das ; Thoudam Doren Singh
COMMENTS:       In Proceedings of the 14th International Workshop on Semantic Evaluation (SemEval-2020), Barcelona, Spain, December. Association for Computational Linguistics
HIGHLIGHT:      This work proposes a system named NITS-Hinglish-SentiMix to viably complete the sentiment analysis of such code-mixed Hinglish text.

4, TITLE:       Pixel-Pair Occlusion Relationship Map(P2ORM): Formulation, Inference & Application
http://arxiv.org/abs/2007.12088
AUTHORS:        Xuchong Qiu ; Yang Xiao ; Chaohui Wang ; Renaud Marlet
COMMENTS:       Accepted to ECCV 2020 as a spotlight. Project page: http://imagine.enpc.fr/~qiux/P2ORM/
HIGHLIGHT:      The former provides a way to generate large-scale accurate occlusion datasets while, based on the latter, we propose a novel method for task-independent pixel-level occlusion relationship estimation from single images.

5, TITLE:       Attention based Multiple Instance Learning for Classification of Blood Cell Disorders
http://arxiv.org/abs/2007.11641
AUTHORS:        Ario Sadafi ; Asya Makhro ; Anna Bogdanova ; Nassir Navab ; Tingying Peng ; Shadi Albarqouni ; Carsten Marr
HIGHLIGHT:      We propose an attention based multiple instance learning method to classify blood samples of patients suffering from blood cell disorders.

6, TITLE:       TSIT: A Simple and Versatile Framework for Image-to-Image Translation
http://arxiv.org/abs/2007.12072
AUTHORS:        Liming Jiang ; Changxu Zhang ; Mingyang Huang ; Chunxiao Liu ; Jianping Shi ; Chen Change Loy
COMMENTS:       ECCV 2020 (Spotlight). GitHub: https://github.com/EndlessSora/TSIT
HIGHLIGHT:      We introduce a simple and versatile framework for image-to-image translation.

7, TITLE:       Representation Sharing for Fast Object Detector Search and Beyond
http://arxiv.org/abs/2007.12075
AUTHORS:        Yujie Zhong ; Zelu Deng ; Sheng Guo ; Matthew R. Scott ; Weilin Huang
COMMENTS:       ECCV 2020 accepted
HIGHLIGHT:      To enhance such capability, we propose an extremely efficient neural architecture search method, named Fast And Diverse (FAD), to better explore the optimal configuration of receptive fields and convolution types in the sub-networks for one-stage detectors.

8, TITLE:       HCMS at SemEval-2020 Task 9: A Neural Approach to Sentiment Analysis for Code-Mixed Texts
http://arxiv.org/abs/2007.12076
AUTHORS:        Aditya Srivastava ; V. Harsha Vardhan
COMMENTS:       6 pages, 2 figures, 4 tables, math equations, to be published in the proceedings of the 14th International Workshop on Semantic Evaluation (SemEval) 2020, Association for Computational Linguistics (ACL). Code for the paper is available at https://github.com/IamAdiSri/hcms-semeval20 . Data and task description is available at https://competitions.codalab.org/competitions/20654
HIGHLIGHT:      In this paper we describe our submission to the Sentimix Hindi-English task involving sentiment classification of code-mixed texts, and with an F1 score of 67.1%, we demonstrate that simple convolution and attention may well produce reasonable results.

9, TITLE:       Polylidar3D -- Fast Polygon Extraction from 3D Data
http://arxiv.org/abs/2007.12065
AUTHORS:        Jeremy Castagno ; Ella Atkins
COMMENTS:       40 pages
HIGHLIGHT:      We present Polylidar3D, a non-convex polygon extraction algorithm which takes as input unorganized 3D point clouds (e.g., LiDAR data), organized point clouds (e.g., range images), or user-provided meshes.

10, TITLE:      Real-time CNN-based Segmentation Architecture for Ball Detection in a Single View Setup
http://arxiv.org/abs/2007.11876
AUTHORS:        Gabriel Van Zandycke ; Christophe De Vleeschouwer
COMMENTS:       8 pages, 10 figures
HIGHLIGHT:      We propose a novel approach by formulating the problem as a segmentation task solved by an efficient CNN architecture. As an additional contribution, we publicly release the dataset on which this work is based.

11, TITLE:      Subjective and Objective Quality Assessment of High Frame Rate Videos
http://arxiv.org/abs/2007.11634
AUTHORS:        Pavan C. Madhusudana ; Xiangxu Yu ; Neil Birkbeck ; Yilin Wang ; Balu Adsumilli ; Alan C. Bovik
HIGHLIGHT:      Towards advancing progression in this direction we designed a new subjective resource, called the LIVE-YouTube-HFR (LIVE-YT-HFR) dataset, which is comprised of 480 videos having 6 different frame rates, obtained from 16 diverse contents.

12, TITLE:      Lower Bounds and Hardness Magnification for Sublinear-Time Shrinking Cellular Automata
http://arxiv.org/abs/2007.12048
AUTHORS:        Augusto Modanese
COMMENTS:       20 pages, 2 figures
HIGHLIGHT:      We prove an equivalent result for the (provably) strictly less capable model of shrinking cellular automata (SCAs), which are cellular automata whose cells can spontaneously delete themselves.

13, TITLE:      ReLaB: Reliable Label Bootstrapping for Semi-Supervised Learning
http://arxiv.org/abs/2007.11866
AUTHORS:        Paul Albert ; Diego Ortego ; Eric Arazo ; Noel E. O'Connor ; Kevin McGuinness
COMMENTS:       8 pages, 3 figures
HIGHLIGHT:      Reducing the amount of labels required to trainconvolutional neural networks without performance degradationis key to effectively reduce human annotation effort.

14, TITLE:      Smooth-AP: Smoothing the Path Towards Large-Scale Image Retrieval
http://arxiv.org/abs/2007.12163
AUTHORS:        Andrew Brown ; Weidi Xie ; Vicky Kalogeiton ; Andrew Zisserman
COMMENTS:       Accepted at ECCV 2020
HIGHLIGHT:      To this end, we introduce an objective that optimises instead a smoothed approximation of AP, coined Smooth-AP.

15, TITLE:      Revealing semantic and emotional structure of suicide notes with cognitive network science
http://arxiv.org/abs/2007.12053
AUTHORS:        Andreia Sofia Teixeira ; Szymon Talaga ; Trevor James Swanson ; Massimo Stella
HIGHLIGHT:      We build upon cognitive network science, psycholinguistics and semantic frame theory to introduce a network representation of the mindset expressed in suicide notes.

16, TITLE:      Learning User-Preferred Mappings for Intuitive Robot Control
http://arxiv.org/abs/2007.11627
AUTHORS:        Mengxi Li ; Dylan P. Losey ; Jeannette Bohg ; Dorsa Sadigh
COMMENTS:       8 pages, 7 figures, Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), October 2020
HIGHLIGHT:      Instead, we propose a personalized method for learning the human's preferred or preconceived mapping from a few robot queries.

17, TITLE:      Health, Psychosocial, and Social issues emanating from COVID-19 pandemic based on Social Media Comments using Natural Language Processing
http://arxiv.org/abs/2007.12144
AUTHORS:        Oladapo Oyebode ; Chinenye Ndulue ; Ashfaq Adib ; Dinesh Mulchandani ; Banuchitra Suruliraj ; Fidelia Anulika Orji ; Christine Chambers ; Sandra Meier ; Rita Orji
HIGHLIGHT:      This paper aims to investigate the impact of the COVID-19 pandemic on people globally using social media data.

18, TITLE:      CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending
http://arxiv.org/abs/2007.12147
AUTHORS:        Hang Xu ; Shaoju Wang ; Xinyue Cai ; Wei Zhang ; Xiaodan Liang ; Zhenguo Li
COMMENTS:       Accepted by ECCV2020
HIGHLIGHT:      In this paper, we propose a novel lane-sensitive architecture search framework named CurveLane-NAS to automatically capture both long-ranged coherent and accurate short-range curve information while unifying both architecture search and post-processing on curve lane predictions via point blending.

19, TITLE:      Spatially Aware Multimodal Transformers for TextVQA
http://arxiv.org/abs/2007.12146
AUTHORS:        Yash Kant ; Dhruv Batra ; Peter Anderson ; Alex Schwing ; Devi Parikh ; Jiasen Lu ; Harsh Agrawal
COMMENTS:       Accepted at European Conference on Computer Vision 2020
HIGHLIGHT:      In contrast, we propose a novel spatially aware self-attention layer such that each visual entity only looks at neighboring entities defined by a spatial graph.

20, TITLE:      Enhanced Transfer Learning for Autonomous Driving with Systematic Accident Simulation
http://arxiv.org/abs/2007.12148
AUTHORS:        Shivam Akhauri ; Laura Zheng ; Ming Lin
COMMENTS:       9 pages; IROS 2020
HIGHLIGHT:      By applying this data to autonomous driving models, we show that transfer learning on simulated data sets provide better generalization and collision avoidance, as compared to random initialization methods.

21, TITLE:      AttentionNAS: Spatiotemporal Attention Cell Search for Video Classification
http://arxiv.org/abs/2007.12034
AUTHORS:        Xiaofang Wang ; Xuehan Xiong ; Maxim Neumann ; AJ Piergiovanni ; Michael S. Ryoo ; Anelia Angelova ; Kris M. Kitani ; Wei Hua
COMMENTS:       ECCV 2020
HIGHLIGHT:      We propose a novel search space for spatiotemporal attention cells, which allows the search algorithm to flexibly explore various design choices in the cell.

22, TITLE:      Privacy-preserving Artificial Intelligence Techniques in Biomedicine
http://arxiv.org/abs/2007.11621
AUTHORS:        Reihaneh Torkzadehmahani ; Reza Nasirigerdeh ; David B. Blumenthal ; Tim Kacprowski ; Markus List ; Julian Matschinske ; Julian Späth ; Nina Kerstin Wenke ; Béla Bihari ; Tobias Frisch ; Anne Hartebrodt ; Anne-Christin Hausschild ; Dominik Heider ; Andreas Holzinger ; Walter Hötzendorfer ; Markus Kastelitz ; Rudolf Mayer ; Cristian Nogales ; Anastasia Pustozerova ; Richard Röttger ; Harald H. H. W. Schmidt ; Ameli Schwalber ; Christof Tschohl ; Andrea Wohner ; Jan Baumbach
COMMENTS:       18 pages, 6 figures, 5 tables
HIGHLIGHT:      This paper provides a structured overview of recent advances in privacy-preserving AI techniques in biomedicine.

23, TITLE:      Differentiable Hierarchical Graph Grouping for Multi-Person Pose Estimation
http://arxiv.org/abs/2007.11864
AUTHORS:        Sheng Jin ; Wentao Liu ; Enze Xie ; Wenhai Wang ; Chen Qian ; Wanli Ouyang ; Ping Luo
COMMENTS:       To appear on ECCV 2020
HIGHLIGHT:      In this paper, we investigate a new perspective of human part grouping and reformulate it as a graph clustering task.

24, TITLE:      Tiny Transfer Learning: Towards Memory-Efficient On-Device Learning
http://arxiv.org/abs/2007.11622
AUTHORS:        Han Cai ; Chuang Gan ; Ligeng Zhu ; Song Han
HIGHLIGHT:      We present Tiny-Transfer-Learning (TinyTL), an efficient on-device learning method to adapt pre-trained models to newly collected data on edge devices.

25, TITLE:      AI4D -- African Language Dataset Challenge
http://arxiv.org/abs/2007.11865
AUTHORS:        Kathleen Siminyu ; Sackey Freshia ; Jade Abbott ; Vukosi Marivate
HIGHLIGHT:      This work details the organisation of the AI4D - African Language Dataset Challenge, an effort to incentivize the creation, organization and discovery of African language datasets through a competitive challenge.

26, TITLE:      Neural Geometric Parser for Single Image Camera Calibration
http://arxiv.org/abs/2007.11855
AUTHORS:        Jinwoo Lee ; Minhyuk Sung ; Hyunjoon Lee ; Junho Kim
COMMENTS:       Accepted for publication at ECCV 2020
HIGHLIGHT:      We propose a neural geometric parser learning single image camera calibration for man-made scenes.

27, TITLE:      Dimension reduction in recurrent networks by canonicalization
http://arxiv.org/abs/2007.12141
AUTHORS:        Lyudmila Grigoryeva ; Juan-Pablo Ortega
COMMENTS:       26 pages
HIGHLIGHT:      The classical notion of canonical state-space realization is adapted in this paper to accommodate semi-infinite inputs so that it can be used as a dimension reduction tool in the recurrent networks setup.

28, TITLE:      HITNet: Hierarchical Iterative Tile Refinement Network for Real-time Stereo Matching
http://arxiv.org/abs/2007.12140
AUTHORS:        Vladimir Tankovich ; Christian Häne ; Sean Fanello ; Yinda Zhang ; Shahram Izadi ; Sofien Bouaziz
HIGHLIGHT:      This paper presents HITNet, a novel neural network architecture for real-time stereo matching.

29, TITLE:      Whole-Body Human Pose Estimation in the Wild
http://arxiv.org/abs/2007.11858
AUTHORS:        Sheng Jin ; Lumin Xu ; Jin Xu ; Can Wang ; Wentao Liu ; Chen Qian ; Wanli Ouyang ; Ping Luo
COMMENTS:       To appear on ECCV2020
HIGHLIGHT:      To fill in this blank, we introduce COCO-WholeBody which extends COCO dataset with whole-body annotations.

30, TITLE:      PIPAL: a Large-Scale Image Quality Assessment Dataset for Perceptual Image Restoration
http://arxiv.org/abs/2007.12142
AUTHORS:        Jinjin Gu ; Haoming Cai ; Haoyu Chen ; Xiaoxing Ye ; Jimmy Ren ; Chao Dong
COMMENTS:       Accepted by ECCV 2020
HIGHLIGHT:      Based on PIPAL, we present new benchmarks for both IQA and super-resolution methods.

31, TITLE:      Discovering Traveling Companions using Autoencoders
http://arxiv.org/abs/2007.11735
AUTHORS:        Xiaochang Li ; Bei Chen ; Xuesong Lu
HIGHLIGHT:      In this work, we propose a generic deep representation learning model using autoencoders, namely, ATTN-MEAN, for the discovery of traveling companions.

32, TITLE:      SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size Sensitive 3D Clothing
http://arxiv.org/abs/2007.11610
AUTHORS:        Garvita Tiwari ; Bharat Lal Bhatnagar ; Tony Tung ; Gerard Pons-Moll
COMMENTS:       European Conference on Computer Vision 2020
HIGHLIGHT:      In this paper, we introduce SizerNet to predict 3D clothing conditioned on human body shape and garment size parameters, and ParserNet to infer garment meshes and shape under clothing with personal details in a single pass from an input mesh. To learn these models, we introduce the SIZER dataset of clothing size variation which includes $100$ different subjects wearing casual clothing items in various sizes, totaling to approximately 2000 scans.

33, TITLE:      The Devil is in Classification: A Simple Framework for Long-tail Instance Segmentation
http://arxiv.org/abs/2007.11978
AUTHORS:        Tao Wang ; Yu Li ; Bingyi Kang ; Junnan Li ; Junhao Liew ; Sheng Tang ; Steven Hoi ; Jiashi Feng
HIGHLIGHT:      This work aims to study and address such open challenges.

34, TITLE:      Sound2Sight: Generating Visual Dynamics from Sound and Context
http://arxiv.org/abs/2007.12130
AUTHORS:        Anoop Cherian ; Moitreya Chatterjee ; Narendra Ahuja
COMMENTS:       Accepted at ECCV 2020
HIGHLIGHT:      In this paper, we study this problem in the context of audio-conditioned visual synthesis -- a task that is important, for example, in occlusion reasoning.

35, TITLE:      BSL-1K: Scaling up co-articulated sign language recognition using mouthing cues
http://arxiv.org/abs/2007.12131
AUTHORS:        Samuel Albanie ; Gül Varol ; Liliane Momeni ; Triantafyllos Afouras ; Joon Son Chung ; Neil Fox ; Andrew Zisserman
COMMENTS:       Appears in: European Conference on Computer Vision 2020 (ECCV 2020). 28 pages
HIGHLIGHT:      In this work, we introduce a new scalable approach to data collection for sign recognition in continuous videos. Finally, (3) we propose new large-scale evaluation sets for the tasks of sign recognition and sign spotting and provide baselines which we hope will serve to stimulate research in this area.

36, TITLE:      Psychological, Computational and Robotic Models of Time Perception
http://arxiv.org/abs/2007.11845
AUTHORS:        Hamit Basgol ; Inci Ayhan ; Emre Ugur
COMMENTS:       49 pages, 6 figures, 1 table
HIGHLIGHT:      This study aims to provide researchers with an interdisciplinary perspective on time.

37, TITLE:      The societal and ethical relevance of computational creativity
http://arxiv.org/abs/2007.11973
AUTHORS:        Michele Loi ; Eleonora Viganò ; Lonneke van der Plas
COMMENTS:       4 pages, 1 figure, Eleventh International Conference on Computational Creativity, ICCC'20
HIGHLIGHT:      In this paper, we provide a philosophical account of the value of creative systems for individuals and society.

38, TITLE:      Comprehensive Image Captioning via Scene Graph Decomposition
http://arxiv.org/abs/2007.11731
AUTHORS:        Yiwu Zhong ; Liwei Wang ; Jianshu Chen ; Dong Yu ; Yin Li
COMMENTS:       Accepted to ECCV 2020
HIGHLIGHT:      We address the challenging problem of image captioning by revisiting the representation of image scene graph.

39, TITLE:      CAD-Deform: Deformable Fitting of CAD Models to 3D Scans
http://arxiv.org/abs/2007.11965
AUTHORS:        Vladislav Ishimtsev ; Alexey Bokhovkin ; Alexey Artemov ; Savva Ignatyev ; Matthias Niessner ; Denis Zorin ; Evgeny Burnaev
COMMENTS:       25 pages, 13 figures, ECCV 2020
HIGHLIGHT:      In this work, we address this shortcoming by introducing CAD-Deform, a method which obtains more accurate CAD-to-scan fits by non-rigidly deforming retrieved CAD models.

40, TITLE:      A weakly supervised registration-based framework for prostate segmentation via the combination of statistical shape model and CNN
http://arxiv.org/abs/2007.11726
AUTHORS:        Chunxia Qin ; Xiaojun Chen ; Jocelyne Troccaz
COMMENTS:       12 pages, 8 figures and 3 tables
HIGHLIGHT:      To address this problem, we proposed a weakly supervised registration-based framework for the precise prostate segmentation, by combining convolutional neural network (CNN) with statistical shape model (SSM).

41, TITLE:      Learning Differentiable Programs with Admissible Neural Heuristics
http://arxiv.org/abs/2007.12101
AUTHORS:        Ameesh Shah ; Eric Zhan ; Jennifer J. Sun ; Abhinav Verma ; Yisong Yue ; Swarat Chaudhuri
COMMENTS:       9 pages, submitted to NeurIPS 2020
HIGHLIGHT:      We study the problem of learning differentiable functions expressed as programs in a domain-specific language.

42, TITLE:      Sign-curing local Hamiltonians: termwise versus global stoquasticity and the use of Clifford transformations
http://arxiv.org/abs/2007.11964
AUTHORS:        Marios Ioannou ; Stephen Piddock ; Milad Marvian ; Joel Klassen ; Barbara M. Terhal
HIGHLIGHT:      We elucidate the distinction between global and termwise stoquasticity for local Hamiltonians and prove several complexity results.

43, TITLE:      Leveraging Bottom-Up and Top-Down Attention for Few-Shot Object Detection
http://arxiv.org/abs/2007.12104
AUTHORS:        Xianyu Chen ; Ming Jiang ; Qi Zhao
COMMENTS:       This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
HIGHLIGHT:      To improve the performance and interpretability of few-shot object detectors, we propose an attentive few-shot object detection network (AttFDNet) that takes the advantages of both top-down and bottom-up attention.

44, TITLE:      Few-Shot Object Detection and Viewpoint Estimation for Objects in the Wild
http://arxiv.org/abs/2007.12107
AUTHORS:        Yang Xiao ; Renaud Marlet
COMMENTS:       Accepted as Poster at ECCV 2020, project website: http://imagine.enpc.fr/~xiaoy/FSDetView/
HIGHLIGHT:      We propose a meta-learning framework that can be applied to both tasks, possibly including 3D data.

45, TITLE:      A Solution to Product detection in Densely Packed Scenes
http://arxiv.org/abs/2007.11946
AUTHORS:        Tianze Rong ; Yanjia Zhu ; Yichao Xiong ; Hongxiang Cai
HIGHLIGHT:      To solve the problem, we proposed a random crop strategy to ensure both the sampling rate and input scale is relatively sufficient as a contrast to the regular random crop.

46, TITLE:      Deep Learning based, end-to-end metaphor detection in Greek language with Recurrent and Convolutional Neural Networks
http://arxiv.org/abs/2007.11949
AUTHORS:        Konstantinos Perifanos ; Eirini Florou ; Dionysis Goutsos
HIGHLIGHT:      This paper presents and benchmarks a number of end-to-end Deep Learning based models for metaphor detection in Greek.

47, TITLE:      Threat of Adversarial Attacks on Face Recognition: A Comprehensive Survey
http://arxiv.org/abs/2007.11709
AUTHORS:        Fatemeh Vakhshiteh ; Raghavendra Ramachandra ; Ahmad Nickabadi
HIGHLIGHT:      In this article, we present a comprehensive survey on adversarial attacks against FR systems and elaborate on the competence of new countermeasures against them.

48, TITLE:      Implicit Latent Variable Model for Scene-Consistent Motion Forecasting
http://arxiv.org/abs/2007.12036
AUTHORS:        Sergio Casas ; Cole Gulino ; Simon Suo ; Katie Luo ; Renjie Liao ; Raquel Urtasun
COMMENTS:       European Conference on Computer Vision (ECCV) 2020
HIGHLIGHT:      In this paper, we aim to learn scene-consistent motion forecasts of complex urban traffic directly from sensor data.

49, TITLE:      Regularization of Building Boundaries in Satellite Images using Adversarial and Regularized Losses
http://arxiv.org/abs/2007.11840
AUTHORS:        Stefano Zorzi ; Friedrich Fraundorfer
HIGHLIGHT:      In this paper we present a method for building boundary refinement and regularization in satellite images using a fully convolutional neural network trained with a combination of adversarial and regularized losses.

50, TITLE:      PClean: Bayesian Data Cleaning at Scale with Domain-Specific Probabilistic Programming
http://arxiv.org/abs/2007.11838
AUTHORS:        Alexander K. Lew ; Monica Agrawal ; David Sontag ; Vikash K. Mansinghka
COMMENTS:       9+8 pages, 4 figures
HIGHLIGHT:      PClean makes three modeling and inference contributions: (i) a domain-general non-parametric generative model of relational data, for inferring latent objects and their network of latent references; (ii) a domain-specific probabilistic programming language, for encoding domain knowledge specific to each dataset being cleaned; and (iii) a domain-general inference engine that adapts to each PClean program by constructing data-driven proposals used in sequential Monte Carlo and particle Gibbs.

51, TITLE:      Groebner basis structure of ideal interpolation
http://arxiv.org/abs/2007.11830
AUTHORS:        Yihe Gong ; Xue Jiang
HIGHLIGHT:      In this paper, we propose the notion of "reverse" complete reduced basis.

52, TITLE:      WeightNet: Revisiting the Design Space of Weight Networks
http://arxiv.org/abs/2007.11823
AUTHORS:        Ningning Ma ; Xiangyu Zhang ; Jiawei Huang ; Jian Sun
COMMENTS:       ECCV 2020
HIGHLIGHT:      We present a conceptually simple, flexible and effective framework for weight generating networks.

53, TITLE:      Funnel Activation for Visual Recognition
http://arxiv.org/abs/2007.11824
AUTHORS:        Ningning Ma ; Xiangyu Zhang ; Jian Sun
COMMENTS:       ECCV 2020
HIGHLIGHT:      We present a conceptually simple but effective funnel activation for image recognition tasks, called Funnel activation (FReLU), that extends ReLU and PReLU to a 2D activation by adding a negligible overhead of spatial condition.

54, TITLE:      Zero-Shot Recognition through Image-Guided Semantic Classification
http://arxiv.org/abs/2007.11814
AUTHORS:        Mei-Chen Yeh ; Fang Li
HIGHLIGHT:      We present a new embedding-based framework for zero-shot learning (ZSL).

55, TITLE:      MuCAN: Multi-Correspondence Aggregation Network for Video Super-Resolution
http://arxiv.org/abs/2007.11803
AUTHORS:        Wenbo Li ; Xin Tao ; Taian Guo ; Lu Qi ; Jiangbo Lu ; Jiaya Jia
COMMENTS:       Accepted By ECCV2020
HIGHLIGHT:      Motivated by these findings, we propose a temporal multi-correspondence aggregation strategy to leverage similar patches across frames, and a cross-scale nonlocal-correspondence aggregation scheme to explore self-similarity of images across scales.

56, TITLE:      Multi-modality imaging with structure-promoting regularisers
http://arxiv.org/abs/2007.11689
AUTHORS:        Matthias J. Ehrhardt
HIGHLIGHT:      In this chapter we discuss mathematical approaches which allow to combine information from several imaging modalities so that multi-modality imaging can be more than just the sum of its components.

57, TITLE:      Autonomous Removal of Perspective Distortion based on Detection Results of Robotic Elevator Button Corner
http://arxiv.org/abs/2007.11806
AUTHORS:        Nachuan Ma
HIGHLIGHT:      In this work, We propose a novel algorithm that can automatically correct perspective distortions of elevator panel images based on button corner detection results.

58, TITLE:      Integrating Image Captioning with Rule-based Entity Masking
http://arxiv.org/abs/2007.11690
AUTHORS:        Aditya Mogadala ; Xiaoyu Shen ; Dietrich Klakow
HIGHLIGHT:      Hence in this paper, we propose a novel framework for the image captioning with an explicit object (e.g., knowledge graph entity) selection process while still maintaining its end-to-end training ability.

59, TITLE:      End-to-End Trainable Deep Active Contour Models for Automated Image Segmentation: Delineating Buildings in Aerial Imagery
http://arxiv.org/abs/2007.11691
AUTHORS:        Ali Hatamizadeh ; Debleena Sengupta ; Demetri Terzopoulos
COMMENTS:       Accepted to European Conference on Computer Vision (ECCV) 2020
HIGHLIGHT:      As a solution, we present Trainable Deep Active Contours (TDACs), an automatic image segmentation framework that intimately unites Convolutional Neural Networks (CNNs) and Active Contour Models (ACMs).

60, TITLE:      Cloud Transformers
http://arxiv.org/abs/2007.11679
AUTHORS:        Kirill Mazur ; Victor Lempitsky
HIGHLIGHT:      We present a new versatile building block for deep point cloud processing architectures.

61, TITLE:      Right for the Right Reason: Making Image Classification Robust
http://arxiv.org/abs/2007.11924
AUTHORS:        Anna Nguyen ; Adrian Oberföll ; Michael Färber
HIGHLIGHT:      In this paper, we propose a new score for automatically quantifying to which degree the model focuses on the right image parts.

62, TITLE:      Contact and Human Dynamics from Monocular Video
http://arxiv.org/abs/2007.11678
AUTHORS:        Davis Rempe ; Leonidas J. Guibas ; Aaron Hertzmann ; Bryan Russell ; Ruben Villegas ; Jimei Yang
COMMENTS:       ECCV 2020
HIGHLIGHT:      In this paper, we present a physics-based method for inferring 3D human motion from video sequences that takes initial 2D and 3D pose estimates as input.

63, TITLE:      Analogical Reasoning for Visually Grounded Language Acquisition
http://arxiv.org/abs/2007.11668
AUTHORS:        Bo Wu ; Haoyu Qin ; Alireza Zareian ; Carl Vondrick ; Shih-Fu Chang
COMMENTS:       12 pages
HIGHLIGHT:      In this paper, we bring this ability to AI, by studying the task of Visually grounded Language Acquisition (VLA).

64, TITLE:      Weakly Supervised 3D Object Detection from Lidar Point Cloud
http://arxiv.org/abs/2007.11901
AUTHORS:        Qinghao Meng ; Wenguan Wang ; Tianfei Zhou ; Jianbing Shen ; Luc Van Gool ; Dengxin Dai
COMMENTS:       ECCV 2020; website: https://github.com/hlesmqh/WS3D
HIGHLIGHT:      This work proposes a weakly supervised approach for 3D object detection, only requiring a small set of weakly annotated scenes, associated with a few precisely labeled object instances.

65, TITLE:      Harnessing spatial homogeneity of neuroimaging data: patch individual filter layers for CNNs
http://arxiv.org/abs/2007.11899
AUTHORS:        Fabian Eitel ; Jan Philipp Albrecht ; Martin Weygandt ; Friedemann Paul ; Kerstin Ritter
HIGHLIGHT:      Here, we suggest a new CNN architecture that combines the idea of hierarchical abstraction in neural networks with a prior on the spatial homogeneity of neuroimaging data: Whereas early layers are trained globally using standard convolutional layers, we introduce for higher, more abstract layers patch individual filters (PIF).

66, TITLE:      Critically Examining the Claimed Value of Convolutions over User-Item Embedding Maps for Recommender Systems
http://arxiv.org/abs/2007.11893
AUTHORS:        Maurizio Ferrari Dacrema ; Federico Parroni ; Paolo Cremonesi ; Dietmar Jannach
HIGHLIGHT:      In this work, we show through analytical considerations and empirical evaluations that the claimed gains reported in the literature cannot be attributed to the ability of CNNs to model embedding correlations, as argued in the original papers.

67, TITLE:      Multi-Compartment Variational Online Learning for Spiking Neural Networks
http://arxiv.org/abs/2007.11894
AUTHORS:        Hyeryung Jang ; Osvaldo Simeone
HIGHLIGHT:      This paper explores a more general model in which each spiking neuron contains multiple compartments, each tracking the dynamics of a distinct membrane potential, while sharing the same synaptic weights across compartments.

68, TITLE:      Darwin's Neural Network: AI-based Strategies for Rapid and Scalable Cell and Coronavirus Screening
http://arxiv.org/abs/2007.11653
AUTHORS:        Sang Won Lee ; Yueh-Ting Chiu ; Philip Brudnicki ; Audrey M. Bischoff ; Angus Jelinek ; Jenny Zijun Wang ; Danielle R. Bogdanowicz ; Andrew F. Laine ; Jia Guo ; Helen H. Lu
COMMENTS:       19 pages, 7 figures
HIGHLIGHT:      Here we adapt the theory of survival of the fittest in the field of computer vision and machine perception to introduce a new framework of multi-class instance segmentation deep learning, Darwin's Neural Network (DNN), to carry out morphometric analysis and classification of COVID19 and MERS-CoV collected in vivo and of multiple mammalian cell types in vitro.

69, TITLE:      Applying GPGPU to Recurrent Neural Network Language Model based Fast Network Search in the Real-Time LVCSR
http://arxiv.org/abs/2007.11794
AUTHORS:        Kyungmin Lee ; Chiyoun Park ; Ilhwan Kim ; Namhoon Kim ; Jaewon Lee
COMMENTS:       4 pages, 2 figures, Interspeech2015(Accepted)
HIGHLIGHT:      This paper proposes a novel method of applying GPGPUs to RNNLM-based graph traversals.

70, TITLE:      SBAT: Video Captioning with Sparse Boundary-Aware Transformer
http://arxiv.org/abs/2007.11888
AUTHORS:        Tao Jin ; Siyu Huang ; Ming Chen ; Yingming Li ; Zhongfei Zhang
COMMENTS:       Appearing at IJCAI 2020
HIGHLIGHT:      In this paper, we focus on the problem of applying the transformer structure to video captioning effectively.

71, TITLE:      End-to-end Learning of Compressible Features
http://arxiv.org/abs/2007.11797
AUTHORS:        Saurabh Singh ; Sami Abu-El-Haija ; Nick Johnston ; Johannes Ballé ; Abhinav Shrivastava ; George Toderici
COMMENTS:       Accepted at ICIP 2020
HIGHLIGHT:      We propose a learned method that jointly optimizes for compressibility along with the task objective for learning the features.

72, TITLE:      Effects of Language Relatedness for Cross-lingual Transfer Learning in Character-Based Language Models
http://arxiv.org/abs/2007.11648
AUTHORS:        Mittul Singh ; Peter Smit ; Sami Virpioja ; Mikko Kurimo
HIGHLIGHT:      In the same vein, we propose to use cross-lingual transfer for character NNLMs applied to low-resource Automatic Speech Recognition (ASR).

73, TITLE:      Accurate RGB-D Salient Object Detection via Collaborative Learning
http://arxiv.org/abs/2007.11782
AUTHORS:        Wei Ji ; Jingjing Li ; Miao Zhang ; Yongri Piao ; Huchuan Lu
COMMENTS:       accepted by ECCV 2020 as a poster
HIGHLIGHT:      In this paper, we propose a novel collaborative learning framework where edge, depth and saliency are leveraged in a more efficient way, which solves those problems tactfully.

74, TITLE:      Dataflow Analysis With Prophecy and History Variables
http://arxiv.org/abs/2007.12015
AUTHORS:        Martin Rinard ; Austin Gadient
HIGHLIGHT:      We present several classical dataflow analyses with this approach (live variables, very busy expressions, defined variables, and reaching definitions) along with proofs that highlight how this approach can enable more streamlined reasoning.

75, TITLE:      Few-shot Visual Reasoning with Meta-analogical Contrastive Learning
http://arxiv.org/abs/2007.12020
AUTHORS:        Youngsung Kim ; Jinwoo Shin ; Eunho Yang ; Sung Ju Hwang
HIGHLIGHT:      In this work, we propose to solve such a few-shot (or low-shot) visual reasoning problem, by resorting to analogical reasoning, which is a unique human ability to identify structural or relational similarity between two sets.

76, TITLE:      Illumination invariant hyperspectral image unmixing based on a digital surface model
http://arxiv.org/abs/2007.11770
AUTHORS:        Tatsumi Uezato ; Naoto Yokoya ; Wei He
HIGHLIGHT:      This paper proposes an unmixing model, named illumination invariant spectral unmixing (IISU).

77, TITLE:      Bridging the Imitation Gap by Adaptive Insubordination
http://arxiv.org/abs/2007.12173
AUTHORS:        Luca Weihs ; Unnat Jain ; Jordi Salvador ; Svetlana Lazebnik ; Aniruddha Kembhavi ; Alexander Schwing
COMMENTS:       The first two authors contributed equally
HIGHLIGHT:      To better address these tasks and alleviate the imitation gap we propose 'Adaptive Insubordination' (ADVISOR), which dynamically reweights imitation and reward-based reinforcement learning losses during training, enabling switching between imitation and exploration.

78, TITLE:      Product Title Generation for Conversational Systems using BERT
http://arxiv.org/abs/2007.11768
AUTHORS:        Mansi Ranjit Mane ; Shashank Kedia ; Aditya Mantha ; Stephen Guo ; Kannan Achan
HIGHLIGHT:      We propose a sequence-to-sequence approach using BERT to generate short, natural, spoken language titles from input web titles.

79, TITLE:      All at Once: Temporally Adaptive Multi-Frame Interpolation with Advanced Motion Modeling
http://arxiv.org/abs/2007.11762
AUTHORS:        Zhixiang Chi ; Rasoul Mohammadi Nasiri ; Zheng Liu ; Juwei Lu ; Jin Tang ; Konstantinos N Plataniotis
COMMENTS:       Accepted at ECCV2020 (poster)
HIGHLIGHT:      Departing from the state-of-the-art, this work introduces a true multi-frame interpolator.

80, TITLE:      Revisiting Locality in Binary-Integer Representations
http://arxiv.org/abs/2007.12159
AUTHORS:        Hrishee Shastri ; Eitan Frachtenberg
HIGHLIGHT:      In this paper we focus on the popular representations of integers from bit strings, and ask how sensitive these particular representations are to variation operators, and in particular to single-bit mutations.

81, TITLE:      Guided Deep Decoder: Unsupervised Image Pair Fusion
http://arxiv.org/abs/2007.11766
AUTHORS:        Tatsumi Uezato ; Danfeng Hong ; Naoto Yokoya ; Wei He
COMMENTS:       ECCV 2020
HIGHLIGHT:      To address this limitation, in this study, we propose a guided deep decoder network as a general prior.

82, TITLE:      Clustering of Social Media Messages for Humanitarian Aid Response during Crisis
http://arxiv.org/abs/2007.11756
AUTHORS:        Swati Padhee ; Tanay Kumar Saha ; Joel Tetreault ; Alejandro Jaimes
COMMENTS:       6 pages, 1 figure. Research work was done while Swati was interning at Dataminr Inc. and presented at the AI for Social Good, Harvard CRCS Workshop 2020 (https://aiforgood2020.github.io)
HIGHLIGHT:      In this work, we show that recent advances in Deep Learning and Natural Language Processing outperform prior approaches for the task of classifying informativeness and encourage the field to adopt them for their research or even deployment.

83, TITLE:      PareCO: Pareto-aware Channel Optimization for Slimmable Neural Networks
http://arxiv.org/abs/2007.11752
AUTHORS:        Ting-Wu Chin ; Ari S. Morcos ; Diana Marculescu
COMMENTS:       preprint, 4-page abridged versions have been accepted at non-archival venues including RealML and DMMLSys workshops at ICML'20 and DLP-KDD and AdvML workshops at KDD'20
HIGHLIGHT:      To allow for heterogeneous width-multipliers across different layers, we formulate the problem of optimizing slimmable networks from a multi-objective optimization lens, which leads to a novel algorithm for optimizing both the shared weights and the width-multipliers for the sub-networks.

84, TITLE:      History Repeats Itself: Human Motion Prediction via Motion Attention
http://arxiv.org/abs/2007.11755
AUTHORS:        Wei Mao ; Miaomiao Liu ; Mathieu Salzmann
COMMENTS:       Accepted by ECCV2020
HIGHLIGHT:      Here, we introduce an attention-based feed-forward network that explicitly leverages this observation.

85, TITLE:      Total Domination in Unit Disk Graphs
http://arxiv.org/abs/2007.11997
AUTHORS:        Sangram K. Jena ; Gautam K. Das
HIGHLIGHT:      Here we consider the TDS problem in unit disk graphs, where the objective is to find a minimum cardinality total dominating set for an input graph.

86, TITLE:      Improving Competence for Reliable Autonomy
http://arxiv.org/abs/2007.11740
AUTHORS:        Connor Basich ; Justin Svegliato ; Kyle Hollins Wray ; Stefan J. Witwicki ; Shlomo Zilberstein
COMMENTS:       In Proceedings AREA 2020, arXiv:2007.11260
HIGHLIGHT:      In this paper, we propose a method for improving the competence of a system over the course of its deployment.

87, TITLE:      Exploratory Experiments on Programming Autonomous Robots in Jadescript
http://arxiv.org/abs/2007.11741
AUTHORS:        Eleonora Iotti ; Giuseppe Petrosino ; Stefania Monica ; Federico Bergenti
COMMENTS:       In Proceedings AREA 2020, arXiv:2007.11260
HIGHLIGHT:      This paper describes exploratory experiments to validate the possibility of programming autonomous robots using an agent-oriented programming language.

88, TITLE:      Adaptable and Verifiable BDI Reasoning
http://arxiv.org/abs/2007.11743
AUTHORS:        Peter Stringer ; Rafael C. Cardoso ; Xiaowei Huang ; Louise A. Dennis
COMMENTS:       In Proceedings AREA 2020, arXiv:2007.11260
HIGHLIGHT:      In this position paper, we describe a system architecture for BDI autonomous agents capable of adapting to changes in a dynamic environment and outline the required research.

89, TITLE:      End-to-End Optimization of Scene Layout
http://arxiv.org/abs/2007.11744
AUTHORS:        Andrew Luo ; Zhoutong Zhang ; Jiajun Wu ; Joshua B. Tenenbaum
COMMENTS:       CVPR 2020 (Oral). Project page: http://3dsln.csail.mit.edu/
HIGHLIGHT:      We propose an end-to-end variational generative model for scene layout synthesis conditioned on scene graphs.

90, TITLE:      The importance of the spectral gap in estimating ground-state energies
http://arxiv.org/abs/2007.11582
AUTHORS:        Abhinav Deshpande ; Alexey V. Gorshkov ; Bill Fefferman
COMMENTS:       32 pages, 4 figures. Comments welcome
HIGHLIGHT:      In this work, we make progress on this question by considering the precise regime, in which one estimates the ground-state energy to within inverse exponential precision.



==========Updates to Previous Papers==========
1, TITLE:       Evolved Neuromorphic Control for High Speed Divergence-based Landings of MAVs
http://arxiv.org/abs/2003.03118
AUTHORS:        J. J. Hagenaars ; F. Paredes-Vallés ; S. M. Bohté ; G. C. H. E. de Croon
COMMENTS:       8 pages, 6 figures, camera-ready version to appear in IEEE Robotics and Automation Letters
HIGHLIGHT:      In light of this, we want to mimic flying insects in terms of their processing capabilities, and consequently show the efficiency of this approach in the real world.

2, TITLE:       Coresets for the Nearest-Neighbor Rule
http://arxiv.org/abs/2002.06650
AUTHORS:        Alejandro Flores-Velazco ; David M. Mount
HIGHLIGHT:      This paper introduces the concept of coresets for nearest-neighbor classification.

3, TITLE:       Learning to learn generative programs with Memoised Wake-Sleep
http://arxiv.org/abs/2007.03132
AUTHORS:        Luke B. Hewitt ; Tuan Anh Le ; Joshua B. Tenenbaum
HIGHLIGHT:      To tackle the challenge of performing program induction as an 'inner-loop' to learning, we propose the Memoised Wake-Sleep (MWS) algorithm, which extends Wake Sleep by explicitly storing and reusing the best programs discovered by the inference network throughout training.

4, TITLE:       Scratch that! An Evolution-based Adversarial Attack against Neural Networks
http://arxiv.org/abs/1912.02316
AUTHORS:        Malhar Jere ; Loris Rossi ; Briland Hitaj ; Gabriela Ciocarlie ; Giacomo Boracchi ; Farinaz Koushanfar
HIGHLIGHT:      We successfully launch our attack against Microsoft's Cognitive Services Image Captioning API and propose various mitigation strategies.

5, TITLE:       Structured Compression and Sharing of Representational Space for Continual Learning
http://arxiv.org/abs/2001.08650
AUTHORS:        Gobinda Saha ; Isha Garg ; Aayush Ankit ; Kaushik Roy
COMMENTS:       17 pages, 10 figures
HIGHLIGHT:      We propose an algorithm that enables a network to learn continually and efficiently by partitioning the learnt space into a Core space, that serves as the condensed knowledge base over previously learned tasks, and a Residual space, which is akin to a scratch space for learning the current task.

6, TITLE:       Parameter estimation for integer-valued Gibbs distributions
http://arxiv.org/abs/1904.03139
AUTHORS:        David G. Harris ; Vladimir Kolmogorov
COMMENTS:       Superseded by arXiv:2007.10824
HIGHLIGHT:      We consider the problem of estimating the normalized coefficients $c_k$ for indices $k\in\cal K$ satisfying $\max_\beta\mu^\Omega_\beta(\{x|H(x)=k\})\ge\mu_*$, where $\mu_*\in(0,1)$ is a given parameter and $\cal K$ is a given subset of $\cal H$.

7, TITLE:       Targeted Attack for Deep Hashing based Retrieval
http://arxiv.org/abs/2004.07955
AUTHORS:        Jiawang Bai ; Bin Chen ; Yiming Li ; Dongxian Wu ; Weiwei Guo ; Shu-tao Xia ; En-hui Yang
COMMENTS:       Accepted by ECCV 2020 as Oral
HIGHLIGHT:      In this paper, we propose a novel method, dubbed deep hashing targeted attack (DHTA), to study the targeted attack on such retrieval.

8, TITLE:       Automatic Analysis System of Calcaneus Radiograph: Rotation-Invariant Landmark Detection for Calcaneal Angle Measurement, Fracture Identification and Fracture Region Segmentation
http://arxiv.org/abs/1912.04536
AUTHORS:        Jia Guo ; Yuxuan Mu ; Huiqi Li ; Junxian Chen ; Wei Wang ; Huanxin Yan ; Hailin Xu
HIGHLIGHT:      The aim of this study is to develop an analysis system that can automatically locate four anatomic landmarks, measure BA and CAG for fracture assessment, identify fractured calcaneus and segment fractured regions.

9, TITLE:       IIT Gandhinagar at SemEval-2020 Task 9: Code-Mixed Sentiment Classification Using Candidate Sentence Generation and Selection
http://arxiv.org/abs/2006.14465
AUTHORS:        Vivek Srivastava ; Mayank Singh
HIGHLIGHT:      We present a candidate sentence generation and selection based approach on top of the Bi-LSTM based neural classifier to classify the Hinglish code-mixed text into one of the three sentiment classes positive, negative, or neutral.

10, TITLE:      Through the Looking Glass: Neural 3D Reconstruction of Transparent Shapes
http://arxiv.org/abs/2004.10904
AUTHORS:        Zhengqin Li ; Yu-Ying Yeh ; Manmohan Chandraker
COMMENTS:       Accepted by CVPR 2020 as an oral presentation
HIGHLIGHT:      Our novel contributions include a normal representation that enables the network to model complex light transport through local computation, a rendering layer that models refractions and reflections, a cost volume specifically designed for normal refinement of transparent shapes and a feature mapping based on predicted normals for 3D point cloud reconstruction.

11, TITLE:      DPDist : Comparing Point Clouds Using Deep Point Cloud Distance
http://arxiv.org/abs/2004.11784
AUTHORS:        Dahlia Urbach ; Yizhak Ben-Shabat ; Michael Lindenbaum
HIGHLIGHT:      We introduce a new deep learning method for point cloud comparison.

12, TITLE:      Deep Learning-based Vehicle Behaviour Prediction For Autonomous Driving Applications: A Review
http://arxiv.org/abs/1912.11676
AUTHORS:        Sajjad Mozaffari ; Omar Y. Al-Jarrah ; Mehrdad Dianati ; Paul Jennings ; Alexandros Mouzakitis
HIGHLIGHT:      Motivated by this increased popularity, we provide a comprehensive review of the state-of-the-art of deep learning-based approaches for vehicle behaviour prediction in this paper.

13, TITLE:      Structured Landmark Detection via Topology-Adapting Deep Graph Learning
http://arxiv.org/abs/2004.08190
AUTHORS:        Weijian Li ; Yuhang Lu ; Kang Zheng ; Haofu Liao ; Chihung Lin ; Jiebo Luo ; Chi-Tung Cheng ; Jing Xiao ; Le Lu ; Chang-Fu Kuo ; Shun Miao
COMMENTS:       Accepted to ECCV-20. Camera-ready with supplementary material
HIGHLIGHT:      In this work, we present a new topology-adapting deep graph learning approach for accurate anatomical facial and medical (e.g., hand, pelvis) landmark detection.

14, TITLE:      A Generalization of Otsu's Method and Minimum Error Thresholding
http://arxiv.org/abs/2007.07350
AUTHORS:        Jonathan T. Barron
COMMENTS:       ECCV 2020
HIGHLIGHT:      We present Generalized Histogram Thresholding (GHT), a simple, fast, and effective technique for histogram-based image thresholding.

15, TITLE:      Resources: A Safe Language Abstraction for Money
http://arxiv.org/abs/2004.05106
AUTHORS:        Sam Blackshear ; David L. Dill ; Shaz Qadeer ; Clark W. Barrett ; John C. Mitchell ; Oded Padon ; Yoni Zohar
HIGHLIGHT:      Addressing this need, we present flexible and reliable abstractions for programming with digital currency in the Move language [Blackshear et al. 2019].

16, TITLE:      Anatomy-Aware Siamese Network: Exploiting Semantic Asymmetry for Accurate Pelvic Fracture Detection in X-ray Images
http://arxiv.org/abs/2007.01464
AUTHORS:        Haomin Chen ; Yirui Wang ; Kang Zheng ; Weijian Li ; Chi-Tung Cheng ; Adam P. Harrison ; Jing Xiao ; Gregory D. Hager ; Le Lu ; Chien-Hung Liao ; Shun Miao
COMMENTS:       ECCV 2020 (camera-ready)
HIGHLIGHT:      In this work, we exploit semantic anatomical symmetry or asymmetry analysis in a complex CAD scenario, i.e., anterior pelvic fracture detection in trauma PXRs, where semantically pathological (refer to as fracture) and non-pathological (e.g., pose) asymmetries both occur.

17, TITLE:      Decidable Synthesis of Programs with Uninterpreted Functions
http://arxiv.org/abs/1910.09744
AUTHORS:        Paul Krogmeier ; Umang Mathur ; Adithya Murali ; P. Madhusudan ; Mahesh Viswanathan
HIGHLIGHT:      We identify a decidable synthesis problem for a class of programs of unbounded size with conditionals and iteration that work over infinite data domains.

18, TITLE:      Discriminative Feature Alignment: Improving Transferability of Unsupervised Domain Adaptation by Gaussian-guided Latent Alignment
http://arxiv.org/abs/2006.12770
AUTHORS:        Jing Wang ; Jiahong Chen ; Jianzhe Lin ; Leonid Sigal ; Clarence W. de Silva
COMMENTS:       15 pages, 12 figures
HIGHLIGHT:      In this study, we focus on the unsupervised domain adaptation problem where an approximate inference model is to be learned from a labeled data domain and expected to generalize well to an unlabeled data domain.

19, TITLE:      Modeling and Enhancing Low-quality Retinal Fundus Images
http://arxiv.org/abs/2005.05594
AUTHORS:        Ziyi Shen ; Huazhu Fu ; Jianbing Shen ; Ling Shao
HIGHLIGHT:      Due to the special optical beam of fundus imaging and retinal structure, the natural image enhancement methods cannot be utilized directly.

20, TITLE:      CrossTransformers: spatially-aware few-shot transfer
http://arxiv.org/abs/2007.11498
AUTHORS:        Carl Doersch ; Ankush Gupta ; Andrew Zisserman
HIGHLIGHT:      In this work, we illustrate how the neural network representations which underpin modern vision systems are subject to supervision collapse, whereby they lose any information that is not necessary for performing the training task, including information that may be necessary for transfer to new tasks or domains.

21, TITLE:      Adversarial Training Reduces Information and Improves Transferability
http://arxiv.org/abs/2007.11259
AUTHORS:        Matteo Terzi ; Alessandro Achille ; Marco Maggipinto ; Gian Antonio Susto
HIGHLIGHT:      The latter property may seem counter-intuitive as it is widely accepted by the community that classification models should only capture the minimal information (features) required for the task.

22, TITLE:      Collaborative Video Object Segmentation by Foreground-Background Integration
http://arxiv.org/abs/2003.08333
AUTHORS:        Zongxin Yang ; Yunchao Wei ; Yi Yang
COMMENTS:       ECCV 2020, Spotlight
HIGHLIGHT:      This paper investigates the principles of embedding learning to tackle the challenging semi-supervised video object segmentation.

23, TITLE:      Interpreting the Latent Space of GANs via Correlation Analysis for Controllable Concept Manipulation
http://arxiv.org/abs/2006.10132
AUTHORS:        Ziqiang Li ; Rentuo Tao ; Hongjing Niu ; Bin Li
HIGHLIGHT:      To get deeper insight of the intrinsic mechanism of GANs, in this paper, a method for interpreting the latent space of GANs by analyzing the correlation between latent variables and the corresponding semantic contents in generated images is proposed.

24, TITLE:      Word Representation for Rhythms
http://arxiv.org/abs/2007.10610
AUTHORS:        Tongyu Lu ; Lucheng Yan ; Gus Xia
COMMENTS:       5 pages, 8 figures Comments: embellished figures; rearranged section 2; added details for figure captions
HIGHLIGHT:      This paper proposes a word representation strategy for rhythm patterns.

25, TITLE:      A Framework for Reinforcement Learning and Planning
http://arxiv.org/abs/2006.15009
AUTHORS:        Thomas M. Moerland ; Joost Broekens ; Catholijn M. Jonker
HIGHLIGHT:      Therefore, this paper presents a unifying framework for reinforcement learning and planning (FRAP), which identifies the underlying dimensions on which any planning or learning algorithm has to decide.

26, TITLE:      Two-Level Attention-based Fusion Learning for RGB-D Face Recognition
http://arxiv.org/abs/2003.00168
AUTHORS:        Hardik Uppal ; Alireza Sepas-Moghaddam ; Michael Greenspan ; Ali Etemad
COMMENTS:       8 Pages, 4 figure, submitted to ICPR2020
HIGHLIGHT:      A novel attention aware method is proposed to fuse two image modalities, RGB and depth, for enhanced RGB-D facial recognition.

27, TITLE:      ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis
http://arxiv.org/abs/2004.01113
AUTHORS:        Eu Wern Teh ; Terrance DeVries ; Graham W. Taylor
COMMENTS:       To appear in the European Conference on Computer Vision (ECCV) 2020
HIGHLIGHT:      We consider the problem of distance metric learning (DML), where the task is to learn an effective similarity measure between images.

28, TITLE:      SelfVIO: Self-Supervised Deep Monocular Visual-Inertial Odometry and Depth Estimation
http://arxiv.org/abs/1911.09968
AUTHORS:        Yasin Almalioglu ; Mehmet Turan ; Alp Eren Sari ; Muhamad Risqi U. Saputra ; Pedro P. B. de Gusmão ; Andrew Markham ; Niki Trigoni
COMMENTS:       15 pages, submitted to The IEEE Transactions on Robotics (T-RO) journal, under review
HIGHLIGHT:      In this study, we introduce a novel self-supervised deep learning-based VIO and depth map recovery approach (SelfVIO) using adversarial training and self-adaptive visual-inertial sensor fusion.

29, TITLE:      Modelling Hierarchical Structure between Dialogue Policy and Natural Language Generator with Option Framework for Task-oriented Dialogue System
http://arxiv.org/abs/2006.06814
AUTHORS:        Jianhong Wang ; Yuan Zhang ; Tae-Kyun Kim ; Yunjie Gu
HIGHLIGHT:      In our work, we (1) propose modelling the hierarchical structure between dialogue policy and natural language generator (NLG) with the option framework, called HDNO; (2) train HDNO with hierarchical reinforcement learning (HRL), as well as suggest alternating updates between dialogue policy and NLG during HRL inspired by fictitious play, to preserve the comprehensibility of generated system utterances while improving fulfilling user requests; and (3) propose using a discriminator modelled with language models as an additional reward to further improve the comprehensibility.

30, TITLE:      Co-occurrence Based Texture Synthesis
http://arxiv.org/abs/2005.08186
AUTHORS:        Anna Darzi ; Itai Lang ; Ashutosh Taklikar ; Hadar Averbuch-Elor ; Shai Avidan
HIGHLIGHT:      In this work, we turn to co-occurrence statistics, which have long been used for texture analysis, to learn a controllable texture synthesis model.

31, TITLE:      TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP
http://arxiv.org/abs/2005.05909
AUTHORS:        John X. Morris ; Eli Lifland ; Jin Yong Yoo ; Jake Grigsby ; Di Jin ; Yanjun Qi
COMMENTS:       6 pages. More details are shared at https://github.com/QData/TextAttack
HIGHLIGHT:      This paper introduces TextAttack, a Python framework for adversarial attacks, data augmentation, and adversarial training in NLP.

32, TITLE:      Deep learning for image segmentation: veritable or overhyped?
http://arxiv.org/abs/1904.08483
AUTHORS:        Zhenzhou Wang
HIGHLIGHT:      This paper gives a short survey of the accuracies achieved by deep learning so far in image classification and image segmentation.

33, TITLE:      Distilling BERT into Simple Neural Networks with Unlabeled Transfer Data
http://arxiv.org/abs/1910.01769
AUTHORS:        Subhabrata Mukherjee ; Ahmed Hassan Awadallah
COMMENTS:       Multilingual version of this work, namely XtremeDistil (https://aka.ms/XtremeDistil) appears at ACL 2020
HIGHLIGHT:      In this work, we leverage large amounts of in-domain unlabeled transfer data in addition to a limited amount of labeled training instances to bridge this gap for distilling BERT.

34, TITLE:      Garment Design with Generative Adversarial Networks
http://arxiv.org/abs/2007.10947
AUTHORS:        Chenxi Yuan ; Mohsen Moghaddam
COMMENTS:       AdvML 2020, KDD workshop
HIGHLIGHT:      This paper explores the capabilities of generative adversarial networks (GAN) for automated attribute-level editing of design concepts.

35, TITLE:      Label-similarity Curriculum Learning
http://arxiv.org/abs/1911.06902
AUTHORS:        Urun Dogan ; Aniket Anand Deshmukh ; Marcin Machura ; Christian Igel
COMMENTS:       Accepted as a conference paper at ECCV 2020
HIGHLIGHT:      We propose a novel curriculum learning approach for image classification that adapts the loss function by changing the label representation.

36, TITLE:      Differentiable Augmentation for Data-Efficient GAN Training
http://arxiv.org/abs/2006.10738
AUTHORS:        Shengyu Zhao ; Zhijian Liu ; Ji Lin ; Jun-Yan Zhu ; Song Han
COMMENTS:       Code: https://github.com/mit-han-lab/data-efficient-gans
HIGHLIGHT:      To combat it, we propose Differentiable Augmentation (DiffAugment), a simple method that improves the data efficiency of GANs by imposing various types of differentiable augmentations on both real and fake samples.

37, TITLE:      Towards Verified Artificial Intelligence
http://arxiv.org/abs/1606.08514
AUTHORS:        Sanjit A. Seshia ; Dorsa Sadigh ; S. Shankar Sastry
HIGHLIGHT:      We describe five challenges for achieving Verified AI, and five corresponding principles for addressing these challenges.

38, TITLE:      Active Visual Information Gathering for Vision-Language Navigation
http://arxiv.org/abs/2007.08037
AUTHORS:        Hanqing Wang ; Wenguan Wang ; Tianmin Shu ; Wei Liang ; Jianbing Shen
COMMENTS:       ECCV2020; website: https://github.com/HanqingWangAI/Active_VLN
HIGHLIGHT:      To achieve this, we propose an end-to-end framework for learning an exploration policy that decides i) when and where to explore, ii) what information is worth gathering during exploration, and iii) how to adjust the navigation decision after the exploration.

39, TITLE:      Abstract Universal Approximation for Neural Networks
http://arxiv.org/abs/2007.06093
AUTHORS:        Zi Wang ; Aws Albarghouthi ; Somesh Jha
HIGHLIGHT:      We present a theoretical result that demonstrates the power of numerical domains, namely, the simple interval domain, for analysis of neural networks.

40, TITLE:      Show, Describe and Conclude: On Exploiting the Structure Information of Chest X-Ray Reports
http://arxiv.org/abs/2004.12274
AUTHORS:        Baoyu Jing ; Zeya Wang ; Eric Xing
COMMENTS:       ACL 2019
HIGHLIGHT:      In this work, we propose a novel framework that exploits the structure information between and within report sections for generating CXR imaging reports.

41, TITLE:      Isotropy Maximization Loss and Entropic Score: Accurate, Fast, Efficient, Scalable, and Turnkey Neural Networks Out-of-Distribution Detection Based on The Principle of Maximum Entropy
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 OOD detection performance of neural networks is due to cross-entropy SoftMax loss anisotropy and extreme propensity to produce low entropy (high confidence) posterior probability distributions in frontal disagreement with the Principle of Maximum Entropy.

42, TITLE:      Why Are Deep Representations Good Perceptual Quality Features?
http://arxiv.org/abs/1812.00412
AUTHORS:        Taimoor Tariq ; Okan Tarhan Tursun ; Munchurl Kim ; Piotr Didyk
COMMENTS:       To be presented at ECCV 2020
HIGHLIGHT:      We introduce two new formulations to measure the frequency and orientation selectivity of the features learned by convolutional layers for evaluating deep features learned by widely-used deep CNNs such as VGG-16.

43, TITLE:      Online Invariance Selection for Local Feature Descriptors
http://arxiv.org/abs/2007.08988
AUTHORS:        Rémi Pautrat ; Viktor Larsson ; Martin R. Oswald ; Marc Pollefeys
COMMENTS:       27 pages, Accepted at ECCV 2020 (Oral)
HIGHLIGHT:      To be invariant, or not to be invariant: that is the question formulated in this work about local descriptors.

44, TITLE:      A Survey of Code-switched Speech and Language Processing
http://arxiv.org/abs/1904.00784
AUTHORS:        Sunayana Sitaram ; Khyathi Raghavi Chandu ; Sai Krishna Rallabandi ; Alan W Black
HIGHLIGHT:      This survey reviews computational approaches for code-switched Speech and Natural Language Processing.

45, TITLE:      Web Similarity in Sets of Search Terms using Database Queries
http://arxiv.org/abs/1502.05957
AUTHORS:        Andrew R. Cohen ; Paul M. B. Vitanyi
COMMENTS:       LaTeX 18 pages, 3 tables. A precursor is arXiv:1308.3177
HIGHLIGHT:      We develop the theory and give applications of classifying using Amazon, Wikipedia, and the NCBI website from the National Institutes of Health.

46, TITLE:      Model-based Reinforcement Learning: A Survey
http://arxiv.org/abs/2006.16712
AUTHORS:        Thomas M. Moerland ; Joost Broekens ; Catholijn M. Jonker
HIGHLIGHT:      This paper presents a survey of the integration of both fields, better known as model-based reinforcement learning.

47, TITLE:      Learning to Match Distributions for Domain Adaptation
http://arxiv.org/abs/2007.10791
AUTHORS:        Chaohui Yu ; Jindong Wang ; Chang Liu ; Tao Qin ; Renjun Xu ; Wenjie Feng ; Yiqiang Chen ; Tie-Yan Liu
COMMENTS:       Preprint. 20 Pages. Code available at https://github.com/jindongwang/transferlearning/tree/master/code/deep/Learning-to-Match
HIGHLIGHT:      This paper proposes Learning to Match (L2M) to automatically learn the cross-domain distribution matching without relying on hand-crafted priors on the matching loss.

48, TITLE:      Dialog Policy Learning for Joint Clarification and Active Learning Queries
http://arxiv.org/abs/2006.05456
AUTHORS:        Aishwarya Padmakumar ; Raymond J. Mooney
COMMENTS:       Corrected errors in Figure 2 and section 4.1
HIGHLIGHT:      In this work, we train a hierarchical dialog policy to jointly perform {\it both} clarification and active learning in the context of an interactive language-based image retrieval task motivated by an on-line shopping application, and demonstrate that jointly learning dialog policies for clarification and active learning is more effective than the use of static dialog policies for one or both of these functions.

49, TITLE:      Semi-DerainGAN: A New Semi-supervised Single Image Deraining Network
http://arxiv.org/abs/2001.08388
AUTHORS:        Yanyan Wei ; Zhao Zhang ; Yang Wang ; Haijun Zhang ; Mingbo Zhao ; Meng Wang
HIGHLIGHT:      In this paper, we propose a new semi-supervised GAN-based deraining network termed Semi-DerainGAN, which can use both synthetic and real rainy images in a uniform network using two supervised and unsupervised processes.

50, TITLE:      Distortion Robust Image Classification using Deep Convolutional Neural Network with Discrete Cosine Transform
http://arxiv.org/abs/1811.05819
AUTHORS:        Md Tahmid Hossain ; Shyh Wei Teng ; Dengsheng Zhang ; Suryani Lim ; Guojun Lu
HIGHLIGHT:      In this work, we propose distortion robust DCT-Net, a Discrete Cosine Transform based module integrated into a deep network which is built on top of VGG16.

51, TITLE:      xCos: An Explainable Cosine Metric for Face Verification Task
http://arxiv.org/abs/2003.05383
AUTHORS:        Yu-Sheng Lin ; Zhe-Yu Liu ; Yu-An Chen ; Yu-Siang Wang ; Hsin-Ying Lee ; Yi-Rong Chen ; Ya-Liang Chang ; Winston H. Hsu
HIGHLIGHT:      In this paper, we propose a novel similarity metric, called explainable cosine ($xCos$), that comes with a learnable module that can be plugged into most of the verification models to provide meaningful explanations.

52, TITLE:      Language Models are Few-Shot Learners
http://arxiv.org/abs/2005.14165
AUTHORS:        Tom B. Brown ; Benjamin Mann ; Nick Ryder ; Melanie Subbiah ; Jared Kaplan ; Prafulla Dhariwal ; Arvind Neelakantan ; Pranav Shyam ; Girish Sastry ; Amanda Askell ; Sandhini Agarwal ; Ariel Herbert-Voss ; Gretchen Krueger ; Tom Henighan ; Rewon Child ; Aditya Ramesh ; Daniel M. Ziegler ; Jeffrey Wu ; Clemens Winter ; Christopher Hesse ; Mark Chen ; Eric Sigler ; Mateusz Litwin ; Scott Gray ; Benjamin Chess ; Jack Clark ; Christopher Berner ; Sam McCandlish ; Alec Radford ; Ilya Sutskever ; Dario Amodei
COMMENTS:       40+32 pages
HIGHLIGHT:      Specifically, we train GPT-3, an autoregressive language model with 175 billion parameters, 10x more than any previous non-sparse language model, and test its performance in the few-shot setting.

53, TITLE:      PointTriNet: Learned Triangulation of 3D Point Sets
http://arxiv.org/abs/2005.02138
AUTHORS:        Nicholas Sharp ; Maks Ovsjanikov
COMMENTS:       21 pages, 9 figures
HIGHLIGHT:      We present PointTriNet, a differentiable and scalable approach enabling point set triangulation as a layer in 3D learning pipelines.

54, TITLE:      Unsupervised Learning with Stein's Unbiased Risk Estimator
http://arxiv.org/abs/1805.10531
AUTHORS:        Christopher A. Metzler ; Ali Mousavi ; Reinhard Heckel ; Richard G. Baraniuk
HIGHLIGHT:      As such, it has seen a flurry of research with new ideas proposed continuously.

55, TITLE:      COVID-19 Twitter Dataset with Latent Topics, Sentiments and Emotions Attributes
http://arxiv.org/abs/2007.06954
AUTHORS:        Raj Kumar Gupta ; Ajay Vishwanath ; Yinping Yang
COMMENTS:       17 pages, 7 figures, 9 tables
HIGHLIGHT:      This resource paper describes a large dataset covering over 63 million coronavirus-related Twitter posts from more than 13 million unique users since 28 January to 1 July 2020.

56, TITLE:      AP20-OLR Challenge: Three Tasks and Their Baselines
http://arxiv.org/abs/2006.03473
AUTHORS:        Zheng Li ; Miao Zhao ; Qingyang Hong ; Lin Li ; Zhiyuan Tang ; Dong Wang ; Liming Song ; Cheng Yang
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:1907.07626, arXiv:1806.00616, arXiv:1706.09742
HIGHLIGHT:      This paper introduces the fifth oriental language recognition (OLR) challenge AP20-OLR, which intends to improve the performance of language recognition systems, along with APSIPA Annual Summit and Conference (APSIPA ASC).

57, TITLE:      VisualCOMET: Reasoning about the Dynamic Context of a Still Image
http://arxiv.org/abs/2004.10796
AUTHORS:        Jae Sung Park ; Chandra Bhagavatula ; Roozbeh Mottaghi ; Ali Farhadi ; Yejin Choi
COMMENTS:       ECCV 2020 Spotlight; Project Website: http://visualcomet.xyz/
HIGHLIGHT:      We propose VisualComet, the novel framework of visual commonsense reasoning tasks to predict events that might have happened before, events that might happen next, and the intents of the people at present.

58, TITLE:      Stochastic Frequency Masking to Improve Super-Resolution and Denoising Networks
http://arxiv.org/abs/2003.07119
AUTHORS:        Majed El Helou ; Ruofan Zhou ; Sabine Süsstrunk
COMMENTS:       ECCV 2020. Project page: https://github.com/majedelhelou/SFM
HIGHLIGHT:      We present an analysis, in the frequency domain, of degradation-kernel overfitting in super-resolution and introduce a conditional learning perspective that extends to both super-resolution and denoising.

59, TITLE:      LionForests: Local Interpretation of Random Forests
http://arxiv.org/abs/1911.08780
AUTHORS:        Ioannis Mollas ; Nick Bassiliades ; Ioannis Vlahavas ; Grigorios Tsoumakas
COMMENTS:       8 Pages, 4 Tables, 6 Figures, Submitted to NeHuAI-2020 Workshop of ECAI2020
HIGHLIGHT:      In this paper, we provide a methodology for shedding light on the predictions of the misjudged family of tree ensemble algorithms.

60, TITLE:      NEMO: Future Object Localization Using Noisy Ego Priors
http://arxiv.org/abs/1909.08150
AUTHORS:        Srikanth Malla ; Isht Dwivedi ; Behzad Dariush ; Chiho Choi
HIGHLIGHT:      To address these problems in a unified approach, we propose NEMO (Noisy Ego MOtion priors for future object localization) for future forecast of agents in the egocentric view.