2020.03.30.txt

==========New Papers==========
1, TITLE:       An improved 3D region detection network: automated detection of the 12th thoracic vertebra in image guided radiation therapy
http://arxiv.org/abs/2003.12163
AUTHORS:        Yunhe Xie ; Gregory Sharp ; David P. Gierga ; Theodore S. Hong ; Thomas Bortfeld ; Kongbin Kang
COMMENTS:       10 pages, 2 figures
HIGHLIGHT:      In this study, we propose a novel 3D full convolutional network (FCN) that is trained to detect anatomical structures from 3D volumetric data, requiring only a small amount of training data.

2, TITLE:       Action Localization through Continual Predictive Learning
http://arxiv.org/abs/2003.12185
AUTHORS:        Sathyanarayanan N. Aakur ; Sudeep Sarkar
COMMENTS:       18 pages, 4 figures and 3 tables
HIGHLIGHT:      In this paper, we present a new approach based on continual learning that uses feature-level predictions for self-supervision.

3, TITLE:       Local Facial Makeup Transfer via Disentangled Representation
http://arxiv.org/abs/2003.12065
AUTHORS:        Zhaoyang Sun ; Wenxuan Liu ; Feng Liu ; Ryan Wen Liu ; Shengwu Xiong
COMMENTS:       15 pages, 10 figures
HIGHLIGHT:      In this paper, we propose a novel unified adversarial disentangling network to further decompose face images into four independent components, i.e., personal identity, lips makeup style, eyes makeup style and face makeup style.

4, TITLE:       HERS: Homomorphically Encrypted Representation Search
http://arxiv.org/abs/2003.12197
AUTHORS:        Joshua J. Engelsma ; Anil K. Jain ; Vishnu Naresh Boddeti
COMMENTS:       25 pages
HIGHLIGHT:      We present a method to search for a probe (or query) image representation against a large gallery in the encrypted domain.

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

6, TITLE:       Augmenting Colonoscopy using Extended and Directional CycleGAN for Lossy Image Translation
http://arxiv.org/abs/2003.12473
AUTHORS:        Shawn Mathew ; Saad Nadeem ; Sruti Kumari ; Arie Kaufman
COMMENTS:       To appear in CVPR 2020. **First two authors contributed equally to this work
HIGHLIGHT:      In this paper, we present a deep learning framework, Extended and Directional CycleGAN, for lossy unpaired image-to-image translation between OC and VC to augment OC video sequences with scale-consistent depth information from VC, and augment VC with patient-specific textures, color and specular highlights from OC (e.g, for realistic polyp synthesis).

7, TITLE:       COVID-19 Screening on Chest X-ray Images Using Deep Learning based Anomaly Detection
http://arxiv.org/abs/2003.12338
AUTHORS:        Jianpeng Zhang ; Yutong Xie ; Yi Li ; Chunhua Shen ; Yong Xia
HIGHLIGHT:      In this work, we aim to develop a new deep anomaly detection model for fast, reliable screening.

8, TITLE:       One-Shot GAN Generated Fake Face Detection
http://arxiv.org/abs/2003.12244
AUTHORS:        Hadi Mansourifar ; Weidong Shi
HIGHLIGHT:      In this paper, we propose a universal One-Shot GAN generated fake face detection method which can be used in significantly different areas of anomaly detection.

9, TITLE:       Towards Discriminability and Diversity: Batch Nuclear-norm Maximization under Label Insufficient Situations
http://arxiv.org/abs/2003.12237
AUTHORS:        Shuhao Cui ; Shuhui Wang ; Junbao Zhuo ; Liang Li ; Qingming Huang ; Qi Tian
COMMENTS:       Accepted to CVPR 2020 as Oral
HIGHLIGHT:      Accordingly, to improve both discriminability and diversity, we propose Batch Nuclear-norm Maximization (BNM) on the output matrix.

10, TITLE:      Multi-Granularity Reference-Aided Attentive Feature Aggregation for Video-based Person Re-identification
http://arxiv.org/abs/2003.12224
AUTHORS:        Zhizheng Zhang ; Cuiling Lan ; Wenjun Zeng ; Zhibo Chen
COMMENTS:       Accepted by CVPR2020
HIGHLIGHT:      In this paper, we propose an attentive feature aggregation module, namely Multi-Granularity Reference-aided Attentive Feature Aggregation (MG-RAFA), to delicately aggregate spatio-temporal features into a discriminative video-level feature representation.

11, TITLE:      Dynamic Region-Aware Convolution
http://arxiv.org/abs/2003.12243
AUTHORS:        Jin Chen ; Xijun Wang ; Zichao Guo ; Xiangyu Zhang ; Jian Sun
COMMENTS:       14 pages
HIGHLIGHT:      We propose a new convolution called Dynamic Region-Aware Convolution (DRConv), which can automatically assign multiple filters to corresponding spatial regions where features have similar representation.

12, TITLE:      Weakly Supervised Dataset Collection for Robust Person Detection
http://arxiv.org/abs/2003.12263
AUTHORS:        Munetaka Minoguchi ; Ken Okayama ; Yutaka Satoh ; Hirokatsu Kataoka
COMMENTS:       Project page: https://github.com/cvpaperchallenge/FashionCultureDataBase_DLoader The paper is under consideration at Pattern Recognition Letters
HIGHLIGHT:      To construct an algorithm that can provide robust person detection, we present a dataset with over 8 million images that was produced in a weakly supervised manner.

13, TITLE:      Learning to Optimize Non-Rigid Tracking
http://arxiv.org/abs/2003.12230
AUTHORS:        Yang Li ; Aljaž Božič ; Tianwei Zhang ; Yanli Ji ; Tatsuya Harada ; Matthias Nießner
COMMENTS:       Accepted to CVPR'2020 (oral)
HIGHLIGHT:      In this paper, we employ learnable optimizations to improve tracking robustness and speed up solver convergence.

14, TITLE:      Learning representations in Bayesian Confidence Propagation neural networks
http://arxiv.org/abs/2003.12415
AUTHORS:        Naresh Balaji Ravichandran ; Anders Lansner ; Pawel Herman
HIGHLIGHT:      In this work we study biologically inspired unsupervised strategies in neural networks based on local Hebbian learning.

15, TITLE:      Controllable Person Image Synthesis with Attribute-Decomposed GAN
http://arxiv.org/abs/2003.12267
AUTHORS:        Yifang Men ; Yiming Mao ; Yuning Jiang ; Wei-Ying Ma ; Zhouhui Lian
COMMENTS:       Accepted by CVPR 2020 (Oral). Project Page: https://menyifang.github.io/projects/ADGAN/ADGAN.html
HIGHLIGHT:      This paper introduces the Attribute-Decomposed GAN, a novel generative model for controllable person image synthesis, which can produce realistic person images with desired human attributes (e.g., pose, head, upper clothes and pants) provided in various source inputs.

16, TITLE:      Towards Accurate Scene Text Recognition with Semantic Reasoning Networks
http://arxiv.org/abs/2003.12294
AUTHORS:        Deli Yu ; Xuan Li ; Chengquan Zhang ; Junyu Han ; Jingtuo Liu ; Errui Ding
COMMENTS:       Accepted to CVPR2020
HIGHLIGHT:      To mitigate these limitations, we propose a novel end-to-end trainable framework named semantic reasoning network (SRN) for accurate scene text recognition, where a global semantic reasoning module (GSRM) is introduced to capture global semantic context through multi-way parallel transmission.

17, TITLE:      Generalizable Semantic Segmentation via Model-agnostic Learning and Target-specific Normalization
http://arxiv.org/abs/2003.12296
AUTHORS:        Jian Zhang ; Lei Qi ; Yinghuan Shi ; Yang Gao
HIGHLIGHT:      To overcome this limitation, we propose a novel domain generalization framework for the generalizable semantic segmentation task, which enhances the generalization ability of the model from two different views, including the training paradigm and the data-distribution discrepancy.

18, TITLE:      CurlingNet: Compositional Learning between Images and Text for Fashion IQ Data
http://arxiv.org/abs/2003.12299
AUTHORS:        Youngjae Yu ; Seunghwan Lee ; Yuncheol Choi ; Yuncheol Choi ; Gunhee Kim
COMMENTS:       4 pages, 4 figures, ICCV 2019 Linguistics Meets image and video retrieval workshop, Fashion IQ challenge
HIGHLIGHT:      We present an approach named CurlingNet that can measure the semantic distance of composition of image-text embedding.

19, TITLE:      A Survey on Edge Intelligence
http://arxiv.org/abs/2003.12172
AUTHORS:        Dianlei Xu ; Tong Li ; Yong Li ; Xiang Su ; Sasu Tarkoma ; Pan Hui
HIGHLIGHT:      In this paper, we present a thorough and comprehensive survey on the literature surrounding edge intelligence.

20, TITLE:      Data-Driven Inference of Representation Invariants
http://arxiv.org/abs/2003.12106
AUTHORS:        Anders Miltner ; Saswat Padhi ; Todd Millstein ; David Walker
COMMENTS:       18 Pages, Full version of PLDI 2020 paper
HIGHLIGHT:      In this paper, we develop a counterexample-driven algorithm for inferring a representation invariant that is sufficient to imply a desired specification for a module.

21, TITLE:      Boolean learning under noise-perturbations in hardware neural networks
http://arxiv.org/abs/2003.12319
AUTHORS:        Louis Andreoli ; Xavier Porte ; Stéphane Chrétien ; Maxime Jacquot ; Laurent Larger ; Daniel Brunner
COMMENTS:       8 pages, 5 figures
HIGHLIGHT:      We find that noise strongly modifies the system's path during convergence, and surprisingly fully decorrelates the final readout weight matrices.

22, TITLE:      Bayesian Hierarchical Multi-Objective Optimization for Vehicle Parking Route Discovery
http://arxiv.org/abs/2003.12508
AUTHORS:        Romit S Beed ; Sunita Sarkar ; Arindam Roy
COMMENTS:       10 pages, 2 Figures, 3 Tables, journal submission
HIGHLIGHT:      This paper proposes a Bayesian hierarchical technique for obtaining the most optimal route to a parking lot.

23, TITLE:      Learning Implicit Surface Light Fields
http://arxiv.org/abs/2003.12406
AUTHORS:        Michael Oechsle ; Michael Niemeyer ; Lars Mescheder ; Thilo Strauss ; Andreas Geiger
HIGHLIGHT:      In this work, we propose a novel implicit representation for capturing the visual appearance of an object in terms of its surface light field.

24, TITLE:      Hybrid Models for Open Set Recognition
http://arxiv.org/abs/2003.12506
AUTHORS:        Hongjie Zhang ; Ang Li ; Jie Guo ; Yanwen Guo
COMMENTS:       17 pages, 5 figures
HIGHLIGHT:      We propose the OpenHybrid framework, which is composed of an encoder to encode the input data into a joint embedding space, a classifier to classify samples to inlier classes, and a flow-based density estimator to detect whether a sample belongs to the unknown category.

25, TITLE:      Weakly-Supervised Action Localization by Generative Attention Modeling
http://arxiv.org/abs/2003.12424
AUTHORS:        Baifeng Shi ; Qi Dai ; Yadong Mu ; Jingdong Wang
COMMENTS:       CVPR2020
HIGHLIGHT:      To solve the problem, in this paper we propose to model the class-agnostic frame-wise probability conditioned on the frame attention using conditional Variational Auto-Encoder (VAE).

26, TITLE:      Assessing Image Quality Issues for Real-World Problem
http://arxiv.org/abs/2003.12511
AUTHORS:        Tai-Yin Chiu ; Yinan Zhao ; Danna Gurari
HIGHLIGHT:      We introduce a new large-scale dataset that links the assessment of image quality issues to two practical vision tasks: image captioning and visual question answering.

27, TITLE:      Enhanced Self-Perception in Mixed Reality: Egocentric Arm Segmentation and Database with Automatic Labelling
http://arxiv.org/abs/2003.12352
AUTHORS:        Ester Gonzalez-Sosa ; Pablo Perez ; Ruben Tolosana ; Redouane Kachach ; Alvaro Villegas
HIGHLIGHT:      In this study, we focus on the egocentric segmentation of arms to improve self-perception in Augmented Virtuality (AV).

28, TITLE:      An Investigation into the Stochasticity of Batch Whitening
http://arxiv.org/abs/2003.12327
AUTHORS:        Lei Huang ; Lei Zhao ; Yi Zhou ; Fan Zhu ; Li Liu ; Ling Shao
COMMENTS:       Accepted to CVPR 2020. The Code is available at https://github.com/huangleiBuaa/StochasticityBW
HIGHLIGHT:      Based on our analysis, we provide a framework for designing and comparing BW algorithms in different scenarios.

29, TITLE:      Lightweight Photometric Stereo for Facial Details Recovery
http://arxiv.org/abs/2003.12307
AUTHORS:        Xueying Wang ; Yudong Guo ; Bailin Deng ; Juyong Zhang
COMMENTS:       Accepted to CVPR2020. The source code is available https://github.com/Juyong/FacePSNet
HIGHLIGHT:      In this paper, we present a lightweight strategy that only requires sparse inputs or even a single image to recover high-fidelity face shapes with images captured under near-field lights. To this end, we construct a dataset containing 84 different subjects with 29 expressions under 3 different lights.

30, TITLE:      Convolutional Spiking Neural Networks for Spatio-Temporal Feature Extraction
http://arxiv.org/abs/2003.12346
AUTHORS:        Ali Samadzadeh ; Fatemeh Sadat Tabatabaei Far ; Ali Javadi ; Ahmad Nickabadi ; Morteza Haghir Chehreghani
COMMENTS:       10 pages, 7 figures, 2 tables
HIGHLIGHT:      In this paper, we provide insight into spatio-temporal feature extraction of convolutional SNNs in experiments designed to exploit this property.

31, TITLE:      Tackling Two Challenges of 6D Object Pose Estimation: Lack of Real Annotated RGB Images and Scalability to Number of Objects
http://arxiv.org/abs/2003.12344
AUTHORS:        Juil Sock ; Pedro Castro ; Anil Armagan ; Guillermo Garcia-Hernando ; Tae-Kyun Kim
HIGHLIGHT:      In this work, we address these two main challenges for 6D object pose estimation and investigate viable methods in experiments.

32, TITLE:      Modeling 3D Shapes by Reinforcement Learning
http://arxiv.org/abs/2003.12397
AUTHORS:        Cheng Lin ; Tingxiang Fan ; Wenping Wang ; Matthias Nießner
COMMENTS:       Video: https://youtu.be/w5e9g_lvbyE
HIGHLIGHT:      Inspired by such artist-based modeling, we propose a two-step neural framework based on RL to learn 3D modeling policies.

33, TITLE:      An Empirical Study of Ownership, Typestate, and Assets in the Obsidian Smart Contract Language
http://arxiv.org/abs/2003.12209
AUTHORS:        Michael Coblenz ; Jonathan Aldrich ; Joshua Sunshine ; Brad Myers
COMMENTS:       22 pages
HIGHLIGHT:      We performed an empirical study with 20 participants comparing Obsidian to Solidity, which is the language most commonly used for writing smart contracts today.

34, TITLE:      Going in circles is the way forward: the role of recurrence in visual inference
http://arxiv.org/abs/2003.12128
AUTHORS:        Ruben S. van Bergen ; Nikolaus Kriegeskorte
HIGHLIGHT:      This important insight suggests that computational neuroscientists may not need to engage recurrent computation, and that computer-vision engineers may be limiting themselves to a special case of FNN if they build recurrent models.

35, TITLE:      Generic bivariate multi-point evaluation, interpolation and modular composition with precomputation
http://arxiv.org/abs/2003.12468
AUTHORS:        Vincent Neiger ; Johan Rosenkilde ; Grigory Solomatov
HIGHLIGHT:      We apply the same technique to modular composition: fix a square-free $G \in \mathbb{K}[x]$ and generic $R \in \mathbb{K}[x]$ both available for precomputation, we then input $f \in \mathbb{K}[x,y]$ and output $f(x, R(x)) ~\mathrm{rem}~ G \in \mathbb{K}[x]$ in quasi-linear time in the size of $f, G, R$.

36, TITLE:      DA-NAS: Data Adapted Pruning for Efficient Neural Architecture Search
http://arxiv.org/abs/2003.12563
AUTHORS:        Xiyang Dai ; Dongdong Chen ; Mengchen Liu ; Yinpeng Chen ; Lu Yuan
HIGHLIGHT:      In this paper, we present DA-NAS that can directly search the architecture for large-scale target tasks while allowing a large candidate set in a more efficient manner.

37, TITLE:      Probabilistic Regression for Visual Tracking
http://arxiv.org/abs/2003.12565
AUTHORS:        Martin Danelljan ; Luc Van Gool ; Radu Timofte
COMMENTS:       CVPR 2020. Includes appendix
HIGHLIGHT:      In this work, we therefore propose a probabilistic regression formulation and apply it to tracking.

38, TITLE:      FFR V1.0: Fon-French Neural Machine Translation
http://arxiv.org/abs/2003.12111
AUTHORS:        Bonaventure F. P. Dossou ; Chris C. Emezue
COMMENTS:       Accepted for the AfricaNLP Workshop, ICLR 2020
HIGHLIGHT:      In this paper, we describe our pilot project: the creation of a large growing corpora for Fon-to-French translations and our FFR v1.0 model, trained on this dataset.

39, TITLE:      Pedestrian Detection with Wearable Cameras for the Blind: A Two-way Perspective
http://arxiv.org/abs/2003.12122
AUTHORS:        Kyungjun Lee ; Daisuke Sato ; Saki Asakawa ; Hernisa Kacorri ; Chieko Asakawa
COMMENTS:       The 2020 ACM CHI Conference on Human Factors in Computing Systems (CHI 2020)
HIGHLIGHT:      We explore this tension from both perspectives, those of sighted passersby and blind users, taking into account camera visibility, in-person versus remote experience, and extracted visual information.

40, TITLE:      Generation of Consistent Sets of Multi-Label Classification Rules with a Multi-Objective Evolutionary Algorithm
http://arxiv.org/abs/2003.12526
AUTHORS:        Thiago Zafalon Miranda ; Diorge Brognara Sardinha ; Márcio Porto Basgalupp ; Yaochu Jin ; Ricardo Cerri
HIGHLIGHT:      In this context, we propose a multi-objective evolutionary algorithm that generates multiple rule-based multi-label classification models, allowing users to choose among models that offer different compromises between predictive power and interpretability.

41, TITLE:      Adversarial System Variant Approximation to Quantify Process Model Generalization
http://arxiv.org/abs/2003.12168
AUTHORS:        Julian Theis ; Houshang Darabi
HIGHLIGHT:      In this paper, a novel deep learning-based methodology called Adversarial System Variant Approximation (AVATAR) is proposed to overcome this issue.

42, TITLE:      Rolling Horizon Evolutionary Algorithms for General Video Game Playing
http://arxiv.org/abs/2003.12331
AUTHORS:        Raluca D. Gaina ; Sam Devlin ; Simon M. Lucas ; Diego Perez-Liebana
HIGHLIGHT:      This paper presents the state of the art in Rolling Horizon Evolutionary algorithms, combining all modifications described in literature and some additional ones for a large resultant hybrid.

43, TITLE:      Identification of Choquet capacity in multicriteria sorting problems through stochastic inverse analysis
http://arxiv.org/abs/2003.12530
AUTHORS:        Renata Pelissari ; Leonardo Tomazeli Duarte
HIGHLIGHT:      We address the problem of Choquet capacity identification for MCSP by applying the Stochastic Acceptability Multicriteri Analysis (SMAA), proposing the SMAA-S-Choquet method.

44, TITLE:      MiLeNAS: Efficient Neural Architecture Search via Mixed-Level Reformulation
http://arxiv.org/abs/2003.12238
AUTHORS:        Chaoyang He ; Haishan Ye ; Li Shen ; Tong Zhang
COMMENTS:       This paper is published in CVPR 2020 (IEEE/CVF Conference on Computer Vision and Pattern Recognition 2020)
HIGHLIGHT:      In this paper, we demonstrate that gradient errors caused by such approximations lead to suboptimality, in the sense that the optimization procedure fails to converge to a (locally) optimal solution.

45, TITLE:      A Distributional Analysis of Sampling-Based Reinforcement Learning Algorithms
http://arxiv.org/abs/2003.12239
AUTHORS:        Philip Amortila ; Doina Precup ; Prakash Panangaden ; Marc G. Bellemare
COMMENTS:       AISTATS 2020
HIGHLIGHT:      We present a distributional approach to theoretical analyses of reinforcement learning algorithms for constant step-sizes.

46, TITLE:      Can you hear me $\textit{now}$? Sensitive comparisons of human and machine perception
http://arxiv.org/abs/2003.12362
AUTHORS:        Michael A Lepori ; Chaz Firestone
COMMENTS:       21 pages; 4 figures
HIGHLIGHT:      Here, we show how this asymmetry can cause such comparisons to underestimate the overlap in human and machine perception.

47, TITLE:      Integrating Crowdsourcing and Active Learning for Classification of Work-Life Events from Tweets
http://arxiv.org/abs/2003.12139
AUTHORS:        Yunpeng Zhao ; Mattia Prosperi ; Tianchen Lyu ; Yi Guo ; Jing Bian
HIGHLIGHT:      To reduce the burden of the manual annotation, yet maintaining its reliability, we devised a crowdsourcing pipeline combined with active learning strategies.

48, TITLE:      Semantic Enrichment of Nigerian Pidgin English for Contextual Sentiment Classification
http://arxiv.org/abs/2003.12450
AUTHORS:        Wuraola Fisayo Oyewusi ; Olubayo Adekanmbi ; Olalekan Akinsande
COMMENTS:       Accepted to ICLR 2020 AfricaNLP workshop
HIGHLIGHT:      By augmenting scarce human labelled code-changed text with ample synthetic code-reformatted text and meaning, we achieve significant improvements in sentiment scoring.

49, TITLE:      Information-Theoretic Probing with Minimum Description Length
http://arxiv.org/abs/2003.12298
AUTHORS:        Elena Voita ; Ivan Titov
HIGHLIGHT:      Instead, we propose an alternative to the standard probes, information-theoretic probing with minimum description length (MDL).

50, TITLE:      Comprehensive Named Entity Recognition on CORD-19 with Distant or Weak Supervision
http://arxiv.org/abs/2003.12218
AUTHORS:        Xuan Wang ; Xiangchen Song ; Yingjun Guan ; Bangzheng Li ; Jiawei Han
HIGHLIGHT:      We also hope this dataset can bring insights for the COVID- 19 studies, both on the biomedical side and on the social side. We created this CORD-19-NER dataset with comprehensive named entity recognition (NER) on the COVID-19 Open Research Dataset Challenge (CORD-19) corpus (2020- 03-13).

51, TITLE:      SaccadeNet: A Fast and Accurate Object Detector
http://arxiv.org/abs/2003.12125
AUTHORS:        Shiyi Lan ; Zhou Ren ; Yi Wu ; Larry S. Davis ; Gang Hua
HIGHLIGHT:      %In this paper, Inspired by such mechanism, we propose a fast and accurate object detector called \textit{SaccadeNet}.

52, TITLE:      Real-time information retrieval from Identity cards
http://arxiv.org/abs/2003.12103
AUTHORS:        Niloofar Tavakolian ; Azadeh Nazemi ; Donal Fitzpatrick
COMMENTS:       6pages,10 figures,conference
HIGHLIGHT:      This paper aims to propose a series of state-of-the-art methods for the journey of an Identification card (ID) from the scanning or capture phase to the point before Optical character recognition (OCR).

53, TITLE:      Cycle Text-To-Image GAN with BERT
http://arxiv.org/abs/2003.12137
AUTHORS:        Trevor Tsue ; Samir Sen ; Jason Li
HIGHLIGHT:      We explore novel approaches to the task of image generation from their respective captions, building on state-of-the-art GAN architectures.

54, TITLE:      AirRL: A Reinforcement Learning Approach to Urban Air Quality Inference
http://arxiv.org/abs/2003.12205
AUTHORS:        Huiqiang Zhong ; Cunxiang Yin ; Xiaohui Wu ; Jinchang Luo ; JiaWei He
HIGHLIGHT:      In this paper, we propose a novel model based on reinforcement learning for urban air quality inference.

55, TITLE:      A Comprehensive Review for Breast Histopathology Image Analysis Using Classical and Deep Neural Networks
http://arxiv.org/abs/2003.12255
AUTHORS:        Xiaomin Zhou ; Chen Li ; Md Mamunur Rahaman ; Yudong Yao ; Shiliang Ai ; Changhao Sun ; Xiaoyan Li ; Qian Wang ; Tao Jiang
HIGHLIGHT:      In this review, we present a comprehensive overview of the BHIA techniques based on ANNs.



==========Updates to Previous Papers==========
1, TITLE:       Central Similarity Quantization for Efficient Image and Video Retrieval
http://arxiv.org/abs/1908.00347
AUTHORS:        Li Yuan ; Tao Wang ; Xiaopeng Zhang ; Zequn Jie ; Francis EH Tay ; Jiashi Feng
COMMENTS:       CVPR2020, Codes: https://github.com/yuanli2333/Hadamard-Matrix-for-hashing
HIGHLIGHT:      In this work, we propose a new \emph{global} similarity metric, termed as \emph{central similarity}, with which the hash codes of similar data pairs are encouraged to approach a common center and those for dissimilar pairs to converge to different centers, to improve hash learning efficiency and retrieval accuracy.

2, TITLE:       Toward Accurate and Realistic Virtual Try-on Through Shape Matching and Multiple Warps
http://arxiv.org/abs/2003.10817
AUTHORS:        Kedan Li ; Min Jin Chong ; Jingen Liu ; David Forsyth
HIGHLIGHT:      A virtual try-on method takes a product image and an image of a model and produces an image of the model wearing the product.

3, TITLE:       Component Analysis for Visual Question Answering Architectures
http://arxiv.org/abs/2002.05104
AUTHORS:        Camila Kolling ; Jônatas Wehrmann ; Rodrigo C. Barros
HIGHLIGHT:      The main goal of this paper is to provide a comprehensive analysis regarding the impact of each component in VQA models.

4, TITLE:       Visual Indeterminacy in GAN Art
http://arxiv.org/abs/1910.04639
AUTHORS:        Aaron Hertzmann
COMMENTS:       Leonardo / SIGGRAPH 2020 Art Papers
HIGHLIGHT:      This paper explores visual indeterminacy as a description for artwork created with Generative Adversarial Networks (GANs).

5, TITLE:       Multi-Agent Task Allocation in Complementary Teams: A Hunter and Gatherer Approach
http://arxiv.org/abs/1912.05748
AUTHORS:        Mehdi Dadvar ; Saeed Moazami ; Harley R. Myler ; Hassan Zargarzadeh
COMMENTS:       15 pages, 12 figures
HIGHLIGHT:      To minimize the collective cost of task accomplishments in a distributed manner, a game-theoretic solution is introduced to couple agents from complementary teams.

6, TITLE:       PanNuke Dataset Extension, Insights and Baselines
http://arxiv.org/abs/2003.10778
AUTHORS:        Jevgenij Gamper ; Navid Alemi Koohbanani ; Simon Graham ; Mostafa Jahanifar ; Syed Ali Khurram ; Ayesha Azam ; Katherine Hewitt ; Nasir Rajpoot
COMMENTS:       Work in progress
HIGHLIGHT:      We study the performance of segmentation and classification models when applied to the proposed dataset and demonstrate the application of models trained on PanNuke to whole-slide images.

7, TITLE:       Estimating Uncertainty and Interpretability in Deep Learning for Coronavirus (COVID-19) Detection
http://arxiv.org/abs/2003.10769
AUTHORS:        Biraja Ghoshal ; Allan Tucker
HIGHLIGHT:      In this paper, we investigate how drop-weights based Bayesian Convolutional Neural Networks (BCNN) can estimate uncertainty in Deep Learning solution to improve the diagnostic performance of the human-machine team using publicly available COVID-19 chest X-ray dataset and show that the uncertainty in prediction is highly correlates with accuracy of prediction.

8, TITLE:       CPR-GCN: Conditional Partial-Residual Graph Convolutional Network in Automated Anatomical Labeling of Coronary Arteries
http://arxiv.org/abs/2003.08560
AUTHORS:        Han Yang ; Xingjian Zhen ; Ying Chi ; Lei Zhang ; Xian-Sheng Hua
COMMENTS:       This work is done by Xingjian Zhen during internship in Alibaba Damo Academy
HIGHLIGHT:      Motivated by the wide application of the graph neural network in structured data, in this paper, we propose a conditional partial-residual graph convolutional network (CPR-GCN), which takes both position and CT image into consideration, since CT image contains abundant information such as branch size and spanning direction.

9, TITLE:       Fast Generation of High Fidelity RGB-D Images by Deep-Learning with Adaptive Convolution
http://arxiv.org/abs/2002.05067
AUTHORS:        Chuhua Xian ; Dongjiu Zhang ; Chengkai Dai ; Charlie C. L. Wang
HIGHLIGHT:      Using the raw data from consumer-level RGB-D cameras as input, we propose a deep-learning based approach to efficiently generate RGB-D images with completed information in high resolution.

10, TITLE:      CNN-based Repetitive self-revised learning for photos' aesthetics imbalanced classification
http://arxiv.org/abs/2003.03081
AUTHORS:        Ying Dai
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:1909.08213
HIGHLIGHT:      In order to realize the auto-assessment of photo aesthetics, we focus on using repetitive self-revised learning (RSRL) to train the CNN-based aesthetics classification network by imbalanced data set.

11, TITLE:      Minimum Class Confusion for Versatile Domain Adaptation
http://arxiv.org/abs/1912.03699
AUTHORS:        Ying Jin ; Ximei Wang ; Mingsheng Long ; Jianmin Wang
HIGHLIGHT:      In this paper, we delve into a missing piece of existing methods: class confusion, the tendency that a classifier confuses the predictions between the correct and ambiguous classes for target examples, which exists in all of the scenarios above.

12, TITLE:      Validation Set Evaluation can be Wrong: An Evaluator-Generator Approach for Maximizing Online Performance of Ranking in E-commerce
http://arxiv.org/abs/2003.11941
AUTHORS:        Guangda Huzhang ; Zhen-Jia Pang ; Yongqing Gao ; Wen-Ji Zhou ; Qing Da ; An-Xiang Zeng ; Yang Yu
HIGHLIGHT:      Validation Set Evaluation can be Wrong: An Evaluator-Generator Approach for Maximizing Online Performance of Ranking in E-commerce

13, TITLE:      P $\approx$ NP, at least in Visual Question Answering
http://arxiv.org/abs/2003.11844
AUTHORS:        Shailza Jolly ; Sebastian Palacio ; Joachim Folz ; Federico Raue ; Joern Hees ; Andreas Dengel
HIGHLIGHT:      In this paper, we measure the potential confounding factors when polar and non-polar samples are used jointly to train a baseline VQA classifier, and compare it to an upper bound where the over-representation of polar questions is excluded from the training.

14, TITLE:      Egoshots, an ego-vision life-logging dataset and semantic fidelity metric to evaluate diversity in image captioning models
http://arxiv.org/abs/2003.11743
AUTHORS:        Pranav Agarwal ; Alejandro Betancourt ; Vana Panagiotou ; Natalia Díaz-Rodríguez
COMMENTS:       15 pages, 25 figures, Accepted at Machine Learning in Real Life (ML-IRL) ICLR 2020 Workshop
HIGHLIGHT:      In this paper, we attempt to show the biased nature of the currently existing image captioning models and present a new image captioning dataset, Egoshots, consisting of 978 real life images with no captions.

15, TITLE:      Kernel Truncated Regression Representation for Robust Subspace Clustering
http://arxiv.org/abs/1705.05108
AUTHORS:        Liangli Zhen ; Dezhong Peng ; Wei Wang ; Xin Yao
COMMENTS:       14 pages
HIGHLIGHT:      To achieve nonlinear subspace clustering, we propose a novel method, called kernel truncated regression representation.

16, TITLE:      MAMNet: Multi-path Adaptive Modulation Network for Image Super-Resolution
http://arxiv.org/abs/1811.12043
AUTHORS:        Jun-Hyuk Kim ; Jun-Ho Choi ; Manri Cheon ; Jong-Seok Lee
HIGHLIGHT:      To address this issue, we propose a novel multi-path adaptive modulation network (MAMNet).

17, TITLE:      Cross-Modality Attention with Semantic Graph Embedding for Multi-Label Classification
http://arxiv.org/abs/1912.07872
AUTHORS:        Renchun You ; Zhiyao Guo ; Lei Cui ; Xiang Long ; Yingze Bao ; Shilei Wen
COMMENTS:       Accepted by AAAI2020
HIGHLIGHT:      Based on the constructed label graph, we propose an adjacency-based similarity graph embedding method to learn semantic label embeddings, which explicitly exploit label relationships.

18, TITLE:      Beyond Notations: Hygienic Macro Expansion for Theorem Proving Languages
http://arxiv.org/abs/2001.10490
AUTHORS:        Sebastian Ullrich ; Leonardo de Moura
COMMENTS:       accepted to IJCAR 2020
HIGHLIGHT:      We take ideas from the Scheme family of programming languages and solve these two problems simultaneously by proposing a novel hygienic macro system custom-built for ITPs.

19, TITLE:      Automatically designing CNN architectures using genetic algorithm for image classification
http://arxiv.org/abs/1808.03818
AUTHORS:        Yanan Sun ; Bing Xue ; Mengjie Zhang ; Gary G. Yen
HIGHLIGHT:      In this paper, we propose an automatic CNN architecture design method by using genetic algorithms, to effectively address the image classification tasks.

20, TITLE:      Evolving Plasticity for Autonomous Learning under Changing Environmental Conditions
http://arxiv.org/abs/1904.01709
AUTHORS:        Anil Yaman ; Giovanni Iacca ; Decebal Constantin Mocanu ; Matt Coler ; George Fletcher ; Mykola Pechenizkiy
HIGHLIGHT:      In this work, we employ genetic algorithms to evolve local learning rules, from Hebbian perspective, to produce autonomous learning under changing environmental conditions.

21, TITLE:      Quantum Distributed Complexity of Set Disjointness on a Line
http://arxiv.org/abs/2002.11795
AUTHORS:        Frederic Magniez ; Ashwin Nayak
COMMENTS:       19 pages, 2 figures. Correction of a typographical error in the statement of Theorem 3.5
HIGHLIGHT:      In this work, we prove an unconditional lower bound of $\widetilde{\Omega}(\sqrt[3]{n d^2}+\sqrt{n} )$ rounds for Set Disjointness on a Line.

22, TITLE:      Conservative set valued fields, automatic differentiation, stochastic gradient method and deep learning
http://arxiv.org/abs/1909.10300
AUTHORS:        Jérôme Bolte ; Edouard Pauwels
HIGHLIGHT:      We introduce generalized derivatives called conservative fields for which we develop a calculus and provide representation formulas.

23, TITLE:      Can we still avoid automatic face detection?
http://arxiv.org/abs/1602.04504
AUTHORS:        Michael J. Wilber ; Vitaly Shmatikov ; Serge Belongie
COMMENTS:       To appear at WACV 2016
HIGHLIGHT:      In this work, we find ways to evade face detection on Facebook, a representative example of a popular social network that uses automatic face detection to enhance their service.

24, TITLE:      Synergic Adversarial Label Learning with DR and AMD for Retinal Image Grading
http://arxiv.org/abs/2003.10607
AUTHORS:        Lie Ju ; Xin Wang ; Xin Zhao ; Paul Bonnington ; Zongyuan Ge
HIGHLIGHT:      We propose a method called synergic adversarial label learning (SALL) which leverages relevant retinal disease labels in both semantic and feature space as additional signals and train the model in a collaborative manner.

25, TITLE:      Visual Commonsense R-CNN
http://arxiv.org/abs/2002.12204
AUTHORS:        Tan Wang ; Jianqiang Huang ; Hanwang Zhang ; Qianru Sun
COMMENTS:       Accepted by CVPR 2020
HIGHLIGHT:      We present a novel unsupervised feature representation learning method, Visual Commonsense Region-based Convolutional Neural Network (VC R-CNN), to serve as an improved visual region encoder for high-level tasks such as captioning and VQA.

26, TITLE:      Learning Multi-Object Tracking and Segmentation from Automatic Annotations
http://arxiv.org/abs/1912.02096
AUTHORS:        Lorenzo Porzi ; Markus Hofinger ; Idoia Ruiz ; Joan Serrat ; Samuel Rota Bulò ; Peter Kontschieder
HIGHLIGHT:      In this work we contribute a novel pipeline to automatically generate training data, and to improve over state-of-the-art multi-object tracking and segmentation (MOTS) methods.

27, TITLE:      Learning to Observe: Approximating Human Perceptual Thresholds for Detection of Suprathreshold Image Transformations
http://arxiv.org/abs/1912.06433
AUTHORS:        Alan Dolhasz ; Carlo Harvey ; Ian Williams
COMMENTS:       8 pages + references
HIGHLIGHT:      In this paper, we propose to directly approximate the perceptual function performed by human observers completing a visual detection task.

28, TITLE:      Cost Volume Pyramid Based Depth Inference for Multi-View Stereo
http://arxiv.org/abs/1912.08329
AUTHORS:        Jiayu Yang ; Wei Mao ; Jose M. Alvarez ; Miaomiao Liu
COMMENTS:       Accepted to 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2020) as Oral Presentation
HIGHLIGHT:      We propose a cost volume-based neural network for depth inference from multi-view images.

29, TITLE:      Three-branch and Mutil-scale learning for Fine-grained Image Recognition (TBMSL-Net)
http://arxiv.org/abs/2003.09150
AUTHORS:        Fan Zhang ; Guisheng Zhai ; Meng Li ; Yizhao Liu
HIGHLIGHT:      Our approach is end-to-end training, through the comprehensive experiments demonstrate that our approach achieves state-of-the-art results on CUB-200-2011, Stanford Cars and FGVC-Aircraft datasets.

30, TITLE:      Gated Channel Transformation for Visual Recognition
http://arxiv.org/abs/1909.11519
AUTHORS:        Zongxin Yang ; Linchao Zhu ; Yu Wu ; Yi Yang
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      In this work, we propose a generally applicable transformation unit for visual recognition with deep convolutional neural networks.

31, TITLE:      Deep Residual Flow for Out of Distribution Detection
http://arxiv.org/abs/2001.05419
AUTHORS:        Ev Zisselman ; Aviv Tamar
HIGHLIGHT:      In this work, we present a novel approach that improves upon the state-of-the-art by leveraging an expressive density model based on normalizing flows.

32, TITLE:      Histogram Layers for Texture Analysis
http://arxiv.org/abs/2001.00215
AUTHORS:        Joshua Peeples ; Weihuang Xu ; Alina Zare
COMMENTS:       16 pages, 6 figures
HIGHLIGHT:      We present a histogram layer for artificial neural networks (ANNs).

33, TITLE:      Retrospective Analysis of the 2019 MineRL Competition on Sample Efficient Reinforcement Learning
http://arxiv.org/abs/2003.05012
AUTHORS:        Stephanie Milani ; Nicholay Topin ; Brandon Houghton ; William H. Guss ; Sharada P. Mohanty ; Keisuke Nakata ; Oriol Vinyals ; Noboru Sean Kuno
COMMENTS:       10 pages, 2 figures
HIGHLIGHT:      We describe the competition and provide an overview of the top solutions, each of which uses deep reinforcement learning and/or imitation learning.

34, TITLE:      Extremely Dense Point Correspondences using a Learned Feature Descriptor
http://arxiv.org/abs/2003.00619
AUTHORS:        Xingtong Liu ; Yiping Zheng ; Benjamin Killeen ; Masaru Ishii ; Gregory D. Hager ; Russell H. Taylor ; Mathias Unberath
COMMENTS:       The work has been accepted for publication in CVPR 2020
HIGHLIGHT:      In this work, we present an effective self-supervised training scheme and novel loss design for dense descriptor learning.

35, TITLE:      Evaluation of Cross-View Matching to Improve Ground Vehicle Localization with Aerial Perception
http://arxiv.org/abs/2003.06515
AUTHORS:        Deeksha Dixit ; Pratap Tokekar
COMMENTS:       8 pages, 10 figures, Submitted to International Conference on Intelligent Robots and Systems (IROS)
HIGHLIGHT:      In this paper, we evaluate cross-view matching for the task of localizing a ground vehicle over a longer trajectory.

36, TITLE:      Splitting Epistemic Logic Programs
http://arxiv.org/abs/1812.08763
AUTHORS:        Pedro Cabalar ; Jorge Fandinno ; Luis Fariñas del Cerro
COMMENTS:       Theory and Practice of Logic Programming
HIGHLIGHT:      In this paper, we propose an extension of the well-known splitting property for logic programs to the epistemic case.

37, TITLE:      Designing Normative Theories of Ethical and Legal Reasoning: LogiKEy Framework, Methodology, and Tool Support
http://arxiv.org/abs/1903.10187
AUTHORS:        Christoph Benzmüller ; Xavier Parent ; Leendert van der Torre
COMMENTS:       50 pages; 10 figures
HIGHLIGHT:      A framework and methodology---termed LogiKEy---for the design and engineering of ethical reasoners, normative theories and deontic logics is presented.

38, TITLE:      Empirical Analysis of Multi-Task Learning for Reducing Model Bias in Toxic Comment Detection
http://arxiv.org/abs/1909.09758
AUTHORS:        Ameya Vaidya ; Feng Mai ; Yue Ning
COMMENTS:       ICWSM 2020
HIGHLIGHT:      We propose a multi-task learning model with an attention layer that jointly learns to predict the toxicity of a comment as well as the identities present in the comments in order to reduce this bias.

39, TITLE:      XTREME: A Massively Multilingual Multi-task Benchmark for Evaluating Cross-lingual Generalization
http://arxiv.org/abs/2003.11080
AUTHORS:        Junjie Hu ; Sebastian Ruder ; Aditya Siddhant ; Graham Neubig ; Orhan Firat ; Melvin Johnson
HIGHLIGHT:      To this end, we introduce the Cross-lingual TRansfer Evaluation of Multilingual Encoders XTREME benchmark, a multi-task benchmark for evaluating the cross-lingual generalization capabilities of multilingual representations across 40 languages and 9 tasks. We release the benchmark to encourage research on cross-lingual learning methods that transfer linguistic knowledge across a diverse and representative set of languages and tasks.

40, TITLE:      Enhancing Out-Of-Domain Utterance Detection with Data Augmentation Based on Word Embeddings
http://arxiv.org/abs/1911.10439
AUTHORS:        Yueqi Feng ; Jiali Lin
COMMENTS:       I see some improvements that can be done. There will be a major change regarding the main idea
HIGHLIGHT:      In this paper, we study how augmented OOD data based on sampling impact OOD utterance detection with a small sample size.

41, TITLE:      Multilingual Culture-Independent Word Analogy Datasets
http://arxiv.org/abs/1911.10038
AUTHORS:        Matej Ulčar ; Kristiina Vaik ; Jessica Lindström ; Milda Dailidėnaitė ; Marko Robnik-Šikonja
COMMENTS:       7 pages, LREC2020 conference
HIGHLIGHT:      We present a collection of such datasets for the word analogy task in nine languages: Croatian, English, Estonian, Finnish, Latvian, Lithuanian, Russian, Slovenian, and Swedish.

42, TITLE:      High Quality ELMo Embeddings for Seven Less-Resourced Languages
http://arxiv.org/abs/1911.10049
AUTHORS:        Matej Ulčar ; Marko Robnik-Šikonja
COMMENTS:       8 pages, 3 figures, LREC2020 conference
HIGHLIGHT:      We offer precomputed embeddings from popular contextual ELMo model for seven languages: Croatian, Estonian, Finnish, Latvian, Lithuanian, Slovenian, and Swedish.

43, TITLE:      Finnish Language Modeling with Deep Transformer Models
http://arxiv.org/abs/2003.11562
AUTHORS:        Abhilash Jain ; Aku Ruohe ; Stig-Arne Grönroos ; Mikko Kurimo
COMMENTS:       4 pages
HIGHLIGHT:      In this project, we investigate the performance of the Transformer architectures-BERT and Transformer-XL for the language modeling task.

44, TITLE:      Relation-Aware Global Attention for Person Re-identification
http://arxiv.org/abs/1904.02998
AUTHORS:        Zhizheng Zhang ; Cuiling Lan ; Wenjun Zeng ; Xin Jin ; Zhibo Chen
COMMENTS:       Accepted by CVPR2020
HIGHLIGHT:      In this work, we propose an effective Relation-Aware Global Attention (RGA) module which captures the global structural information for better attention learning.

45, TITLE:      Additive Angular Margin for Few Shot Learning to Classify Clinical Endoscopy Images
http://arxiv.org/abs/2003.10033
AUTHORS:        Sharib Ali ; Binod Bhattarai ; Tae-Kyun Kim ; Jens Rittscher
COMMENTS:       10 pages
HIGHLIGHT:      In this work, we propose to use a few-shot learning approach that requires less training data and can be used to predict label classes of test samples from an unseen dataset.

46, TITLE:      Generative Adversarial Networks And Domain Adaptation For Training Data Independent Image Registration
http://arxiv.org/abs/1910.08593
AUTHORS:        Dwarikanath Mahapatra
HIGHLIGHT:      We present a DL based approach that can perform medical image registration of one image type despite being trained with images of a different type.