2020.05.21.txt

==========New Papers==========
1, TITLE:       FashionBERT: Text and Image Matching with Adaptive Loss for Cross-modal Retrieval
http://arxiv.org/abs/2005.09801
AUTHORS:        Dehong Gao ; Linbo Jin ; Ben Chen ; Minghui Qiu ; Yi Wei ; Yi Hu ; Hao Wang
COMMENTS:       10 pages, to be published in SIGIR20 Industry Track
HIGHLIGHT:      In this paper, we address the text and image matching in cross-modal retrieval of the fashion industry.

2, TITLE:       Active Speakers in Context
http://arxiv.org/abs/2005.09812
AUTHORS:        Juan Leon Alcazar ; Fabian Caba Heilbron ; Long Mai ; Federico Perazzi ; Joon-Young Lee ; Pablo Arbelaez ; Bernard Ghanem
HIGHLIGHT:      This paper introduces the Active Speaker Context, a novel representation that models relationships between multiple speakers over long time horizons.

3, TITLE:       A Computational Analysis of Polarization on Indian and Pakistani Social Media
http://arxiv.org/abs/2005.09803
AUTHORS:        Aman Tyagi ; Anjalie Field ; Priyank Lathwal ; Yulia Tsvetkov ; Kathleen M. Carley
HIGHLIGHT:      In this work, we examine polarizing messaging on Twitter during these events, particularly focusing on the positions of Indian and Pakistani politicians.

4, TITLE:       PyChain: A Fully Parallelized PyTorch Implementation of LF-MMI for End-to-End ASR
http://arxiv.org/abs/2005.09824
AUTHORS:        Yiwen Shao ; Yiming Wang ; Daniel Povey ; Sanjeev Khudanpur
COMMENTS:       Submtted to Interspeech 2020
HIGHLIGHT:      We present PyChain, a fully parallelized PyTorch implementation of end-to-end lattice-free maximum mutual information (LF-MMI) training for the so-called \emph{chain models} in the Kaldi automatic speech recognition (ASR) toolkit.

5, TITLE:       Mirror Descent Policy Optimization
http://arxiv.org/abs/2005.09814
AUTHORS:        Manan Tomar ; Lior Shani ; Yonathan Efroni ; Mohammad Ghavamzadeh
HIGHLIGHT:      We propose deep Reinforcement Learning (RL) algorithms inspired by mirror descent, a well-known first-order trust region optimization method for solving constrained convex problems.

6, TITLE:       Relevant Region Prediction for Crowd Counting
http://arxiv.org/abs/2005.09816
AUTHORS:        Xinya Chen ; Yanrui Bin ; Changxin Gao ; Nong Sang ; Hao Tang
COMMENTS:       accepted by Neurocomputing
HIGHLIGHT:      In this paper, we propose Relevant Region Prediction (RRP) for crowd counting, which consists of the Count Map and the Region Relation-Aware Module (RRAM).

7, TITLE:       Deep Learning for LiDAR Point Clouds in Autonomous Driving: A Review
http://arxiv.org/abs/2005.09830
AUTHORS:        Ying Li ; Lingfei Ma ; Zilong Zhong ; Fei Liu ; Dongpu Cao ; Jonathan Li ; Michael A. Chapman
COMMENTS:       21 pages, submitted to IEEE Transactions on Neural Networks and Learning Systems
HIGHLIGHT:      In this paper, we provide a systematic review of existing compelling deep learning architectures applied in LiDAR point clouds, detailing for specific tasks in autonomous driving such as segmentation, detection, and classification.

8, TITLE:       Learning and Reasoning for Robot Dialog and Navigation Tasks
http://arxiv.org/abs/2005.09833
AUTHORS:        Keting Lu ; Shiqi Zhang ; Peter Stone ; Xiaoping Chen
COMMENTS:       Accepted to SIGDIAL 2020. arXiv admin note: substantial text overlap with arXiv:1809.11074
HIGHLIGHT:      In this research, we develop algorithms for robot task completions, while looking into the complementary strengths of reinforcement learning and probabilistic reasoning techniques.

9, TITLE:       Attention-based network for low-light image enhancement
http://arxiv.org/abs/2005.09829
AUTHORS:        Cheng Zhang ; Qingsen Yan ; Yu zhu ; Jinqiu Sun ; Yanning Zhang
HIGHLIGHT:      To address such a difficult problem, this paper presents a novel attention-based neural network to generate high-quality enhanced low-light images from the raw sensor data.

10, TITLE:      Positive emotions help rank negative reviews in e-commerce
http://arxiv.org/abs/2005.09837
AUTHORS:        Di Weng ; Jichang Zhao
COMMENTS:       Emotion lexicons are publicly available at https://doi.org/10.6084/m9.figshare.12327680.v1
HIGHLIGHT:      In this paper, we also enrich the previous understandings of emotions in valuing reviews.

11, TITLE:      Computations and Complexities of Tarski's Fixed Points and Supermodular Games
http://arxiv.org/abs/2005.09836
AUTHORS:        Chuangyin Dang ; Qi Qi ; Yinyu Ye
HIGHLIGHT:      We consider two models of computation for Tarski's order preserving function f related to fixed points in a complete lattice: the oracle function model and the polynomial function model.

12, TITLE:      A Further Study of Unsupervised Pre-training for Transformer Based Speech Recognition
http://arxiv.org/abs/2005.09862
AUTHORS:        Dongwei Jiang ; Wubo Li ; Ruixiong Zhang ; Miao Cao ; Ne Luo ; Yang Han ; Wei Zou ; Xiangang Li
HIGHLIGHT:      In this paper, we conduct a further study on MPC and focus on three important aspects: the effect of pre-training data speaking style, its extension on streaming model, and how to better transfer learned knowledge from pre-training stage to downstream tasks.

13, TITLE:      A reinforcement learning based decision support system in textile manufacturing process
http://arxiv.org/abs/2005.09867
AUTHORS:        Zhenglei He ; Kim Phuc Tran ; Sébastien Thomassey ; Xianyi Zeng ; Changhai Yi
HIGHLIGHT:      This paper introduced a reinforcement learning based decision support system in textile manufacturing process.

14, TITLE:      Automated Copper Alloy Grain Size Evaluation Using a Deep-learning CNN
http://arxiv.org/abs/2005.09634
AUTHORS:        George S. Baggs ; Paul Guerrier ; Andrew Loeb ; Jason C. Jones
HIGHLIGHT:      Automated Copper Alloy Grain Size Evaluation Using a Deep-learning CNN

15, TITLE:      Interactive exploration of population scale pharmacoepidemiology datasets
http://arxiv.org/abs/2005.09890
AUTHORS:        Tengel Ekrem Skar ; Einar Holsbø ; Kristian Svendsen ; Lars Ailo Bongo
HIGHLIGHT:      We use Spark to preprocess the data for machine learning and for analyses using SQL queries.

16, TITLE:      Representation Learning with Fine-grained Patterns
http://arxiv.org/abs/2005.09681
AUTHORS:        Yuanhong Xu ; Qi Qian ; Hao Li ; Rong Jin ; Juhua Hu
HIGHLIGHT:      In this work, we investigate a prevalent problem in real-world applications, where the training set only accesses to the supervised information from superclasses but the target task is defined on fine-grained classes.

17, TITLE:      Exploring Transformers for Large-Scale Speech Recognition
http://arxiv.org/abs/2005.09684
AUTHORS:        Liang Lu ; Changliang Liu ; Jinyu Li ; Yifan Gong
COMMENTS:       5 pages, 1 figure
HIGHLIGHT:      In this paper, we aim at understanding the behaviors of Transformers in the large-scale speech recognition setting, where we have used around 65,000 hours of training data.

18, TITLE:      BERTweet: A pre-trained language model for English Tweets
http://arxiv.org/abs/2005.10200
AUTHORS:        Dat Quoc Nguyen ; Thanh Vu ; Anh Tuan Nguyen
HIGHLIGHT:      We present BERTweet, the first public large-scale pre-trained language model for English Tweets.

19, TITLE:      Examining the State-of-the-Art in News Timeline Summarization
http://arxiv.org/abs/2005.10107
AUTHORS:        Demian Gholipour Ghalandari ; Georgiana Ifrim
COMMENTS:       Camera-ready version for ACL 2020
HIGHLIGHT:      In this paper, we compare different TLS strategies using appropriate evaluation frameworks, and propose a simple and effective combination of methods that improves over the state-of-the-art on all tested benchmarks. For a more robust evaluation, we also present a new TLS dataset, which is larger and spans longer time periods than previous datasets.

20, TITLE:      A Comparison of Label-Synchronous and Frame-Synchronous End-to-End Models for Speech Recognition
http://arxiv.org/abs/2005.10113
AUTHORS:        Linhao Dong ; Cheng Yi ; Jianzong Wang ; Shiyu Zhou ; Shuang Xu ; Xueli Jia ; Bo Xu
COMMENTS:       4 pages, 2 figures
HIGHLIGHT:      In this work, we make a detailed comparison on a representative label-synchronous model (transformer) and a soft frame-synchronous model (continuous integrate-and-fire (CIF) based model).

21, TITLE:      Combining the Causal Judgments of Experts with Possibly Different Focus Areas
http://arxiv.org/abs/2005.10131
AUTHORS:        Meir Friedenberg ; Joseph Y. Halpern
COMMENTS:       Appear in the Proceedings of the Sixteenth International Conference on Principles of Knowledge Representation and Reasoning (KR2018}, 2018
HIGHLIGHT:      In this work we show how causal models can be combined in cases where the experts might disagree on the causal structure for variables that appear in both models due to having different focus areas.

22, TITLE:      Discriminative Dictionary Design for Action Classification in Still Images and Videos
http://arxiv.org/abs/2005.10149
AUTHORS:        Abhinaba Roy ; Biplab Banerjee
HIGHLIGHT:      In this paper, we address the problem of action recognition from still images and videos.

23, TITLE:      The Iteration Number of Colour Refinement
http://arxiv.org/abs/2005.10182
AUTHORS:        Sandra Kiefer ; Brendan D. McKay
COMMENTS:       22 pages, 3 figures, full version of a paper accepted at ICALP 2020
HIGHLIGHT:      Modifying the infinite families that we present, we show that for every natural number n >= 10, there are graphs on n vertices on which Colour Refinement requires at least n-2 iterations to reach stabilisation.

24, TITLE:      Combining Experts' Causal Judgments
http://arxiv.org/abs/2005.10180
AUTHORS:        Dalal Alrajeh ; Hana Chockler ; Joseph Y. Halpern
COMMENTS:       A preliminary version of the paper appeared in \emph{Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18)}, 2018}
HIGHLIGHT:      We define a notion of two causal models being \emph{compatible}, and show how compatible causal models can be merged.

25, TITLE:      Feature Purification: How Adversarial Training Performs Robust Deep Learning
http://arxiv.org/abs/2005.10190
AUTHORS:        Zeyuan Allen-Zhu ; Yuanzhi Li
HIGHLIGHT:      In this paper, we present a principle that we call "feature purification", where we show the existence of adversarial examples are due to the accumulation of certain "dense mixtures" in the hidden weights during the training process of a neural network; and more importantly, one of the goals of adversarial training is to remove such mixtures to "purify" hidden weights.

26, TITLE:      End-to-End Speaker Diarization for an Unknown Number of Speakers with Encoder-Decoder Based Attractors
http://arxiv.org/abs/2005.09921
AUTHORS:        Shota Horiguchi ; Yusuke Fujita ; Shinji Watanabe ; Yawen Xue ; Kenji Nagamatsu
HIGHLIGHT:      End-to-end speaker diarization for an unknown number of speakers is addressed in this paper.

27, TITLE:      Fast Decoding of Codes in the Rank, Subspace, and Sum-Rank Metric
http://arxiv.org/abs/2005.09916
AUTHORS:        Hannes Bartz ; Thomas Jerkovits ; Sven Puchinger ; Johan Rosenkilde
HIGHLIGHT:      To accomplish this, we describe a skew-analogue of the existing PM-Basis algorithm for matrices over usual polynomials.

28, TITLE:      Rethinking Performance Estimation in Neural Architecture Search
http://arxiv.org/abs/2005.09917
AUTHORS:        Xiawu Zheng ; Rongrong Ji ; Qiang Wang ; Qixiang Ye ; Zhenguo Li ; Yonghong Tian ; Qi Tian
HIGHLIGHT:      In this paper, we provide a novel yet systematic rethinking of PE in a resource constrained regime, termed budgeted PE (BPE), which precisely and effectively estimates the performance of an architecture sampled from an architecture space.

29, TITLE:      Range Conditioned Dilated Convolutions for Scale Invariant 3D Object Detection
http://arxiv.org/abs/2005.09927
AUTHORS:        Alex Bewley ; Pei Sun ; Thomas Mensink ; Dragomir Anguelov ; Cristian Sminchisescu
COMMENTS:       20 pages
HIGHLIGHT:      This paper presents a novel 3D object detection framework that processes LiDAR data directly on a representation of the sensor's native range images.

30, TITLE:      Relative Positional Encoding for Speech Recognition and Direct Translation
http://arxiv.org/abs/2005.09940
AUTHORS:        Ngoc-Quan Pham ; Thanh-Le Ha ; Tuan-Nam Nguyen ; Thai-Son Nguyen ; Elizabeth Salesky ; Sebastian Stueker ; Jan Niehues ; Alexander Waibel
COMMENTS:       Submitted to Interspeech 2020
HIGHLIGHT:      In this work, we adapt the relative position encoding scheme to the Speech Transformer, where the key addition is relative distance between input states in the self-attention network.

31, TITLE:      GM-CTSC at SemEval-2020 Task 1: Gaussian Mixtures Cross Temporal Similarity Clustering
http://arxiv.org/abs/2005.09946
AUTHORS:        Pierluigi Cassotti ; Annalina Caputo ; Marco Polignano ; Pierpaolo Basile
HIGHLIGHT:      This paper describes the system proposed for the SemEval-2020 Task 1: Unsupervised Lexical Semantic Change Detection.

32, TITLE:      Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting
http://arxiv.org/abs/2005.09704
AUTHORS:        Zili Yi ; Qiang Tang ; Shekoofeh Azizi ; Daesik Jang ; Zhan Xu
COMMENTS:       CVPR 2020 oral paper. 22 pages, 11 figures
HIGHLIGHT:      Motivated by this, we propose a Contextual Residual Aggregation (CRA) mechanism that can produce high-frequency residuals for missing contents by weighted aggregating residuals from contextual patches, thus only requiring a low-resolution prediction from the network.

33, TITLE:      Monte Carlo Inverse Folding
http://arxiv.org/abs/2005.09961
AUTHORS:        Tristan Cazenave ; Thomas Fournier
HIGHLIGHT:      We propose to adapt and evaluate different Monte Carlo Search algorithms for the RNA Inverse Folding problem.

34, TITLE:      Iterative Network for Image Super-Resolution
http://arxiv.org/abs/2005.09964
AUTHORS:        Yuqing Liu ; Shiqi Wang ; Jian Zhang ; Shanshe Wang ; Siwei Ma ; Wen Gao
COMMENTS:       12 pages, 14 figures
HIGHLIGHT:      This paper proposes a substantially different approach relying on the iterative optimization on HR space with an iterative super-resolution network (ISRN).

35, TITLE:      Inverse problems with second-order Total Generalized Variation constraints
http://arxiv.org/abs/2005.09725
AUTHORS:        Kristian Bredies ; Tuomo Valkonen
COMMENTS:       Published in 2011 as a conference proceeding. Uploaded in 2020 on arXiv to ensure availability: the original proceedings are no longer online
HIGHLIGHT:      Inverse problems with second-order Total Generalized Variation constraints

36, TITLE:      Hidden Markov Models and their Application for Predicting Failure Events
http://arxiv.org/abs/2005.09971
AUTHORS:        Paul Hofmann ; Zaid Tashman
COMMENTS:       Will be published in the proceedings of ICCS 2020; @Booklet{EasyChair:3183, author = {Paul Hofmann and Zaid Tashman}, title = {Hidden Markov Models and their Application for Predicting Failure Events}, howpublished = {EasyChair Preprint no. 3183}, year = {EasyChair, 2020}}
HIGHLIGHT:      We show how Markov mixed membership models (MMMM) can be used to predict the degradation of assets.

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

38, TITLE:      AutoML Segmentation for 3D Medical Image Data: Contribution to the MSD Challenge 2018
http://arxiv.org/abs/2005.09978
AUTHORS:        Oliver Rippel ; Leon Weninger ; Dorit Merhof
HIGHLIGHT:      Such an algorithm was developed and is presented in this paper.

39, TITLE:      Creative Artificial Intelligence -- Algorithms vs. humans in an incentivized writing competition
http://arxiv.org/abs/2005.09980
AUTHORS:        Nils Köbis ; Luca Mossink
HIGHLIGHT:      We discuss what these results convey about the performance of NLG algorithms to produce human-like text and propose methodologies to study such learning algorithms in experimental settings.

40, TITLE:      GLEAKE: Global and Local Embedding Automatic Keyphrase Extraction
http://arxiv.org/abs/2005.09740
AUTHORS:        Javad Rafiei Asl ; Juan M. Banda
HIGHLIGHT:      In this work, we introduce Global and Local Embedding Automatic Keyphrase Extractor (GLEAKE) for the task of AKE.

41, TITLE:      On Evaluating Weakly Supervised Action Segmentation Methods
http://arxiv.org/abs/2005.09743
AUTHORS:        Yaser Souri ; Alexander Richard ; Luca Minciullo ; Juergen Gall
COMMENTS:       Technical Report
HIGHLIGHT:      In this work, we focus on two aspects of the use and evaluation of weakly supervised action segmentation approaches that are often overlooked: the performance variance over multiple training runs and the impact of selecting feature extractors for this task.To tackle the first problem, we train each method on the Breakfast dataset 5 times and provide average and standard deviation of the results.

42, TITLE:      A Modified Fourier-Mellin Approach for Source Device Identification on Stabilized Videos
http://arxiv.org/abs/2005.09984
AUTHORS:        Sara Mandelli ; Fabrizio Argenti ; Paolo Bestagini ; Massimo Iuliani ; Alessandro Piva ; Stefano Tubaro
HIGHLIGHT:      In this paper, we propose to overcome this limitation by searching scaling and rotation parameters in the frequency domain.

43, TITLE:      Adapting a Kidney Exchange Algorithm to Align with Human Values
http://arxiv.org/abs/2005.09755
AUTHORS:        Rachel Freedman ; Jana Schaich Borg ; Walter Sinnott-Armstrong ; John P. Dickerson ; Vincent Conitzer
HIGHLIGHT:      In this paper, we provide an end-to-end methodology for estimating weights of individual participant profiles in a kidney exchange.

44, TITLE:      Self-supervised Dynamic CT Perfusion Image Denoising with Deep Neural Networks
http://arxiv.org/abs/2005.09766
AUTHORS:        Dufan Wu ; Hui Ren ; Quanzheng Li
COMMENTS:       13 pages, 9 figures
HIGHLIGHT:      In this paper, we proposed a self-supervised deep learning method for CTP denoising, which did not require any high-dose reference images for training.

45, TITLE:      Reducing Overlearning through Disentangled Representations by Suppressing Unknown Tasks
http://arxiv.org/abs/2005.10220
AUTHORS:        Naveen Panwar ; Tarun Tater ; Anush Sankaran ; Senthil Mani
COMMENTS:       Added appendix with additional results
HIGHLIGHT:      In this research, we propose a three-fold novel contribution: (i) a model-agnostic solution for reducing model overlearning by suppressing all the unknown tasks, (ii) a novel metric to measure the trust score of a trained deep learning model, and (iii) a simulated benchmark dataset, PreserveTask, having five different fundamental image classification tasks to study the generalization nature of models.

46, TITLE:      BlaBla: Linguistic Feature Extraction for Clinical Analysis in Multiple Languages
http://arxiv.org/abs/2005.10219
AUTHORS:        Abhishek Shivkumar ; Jack Weston ; Raphael Lenain ; Emil Fristed
COMMENTS:       5 pages. 1 figure. Under review
HIGHLIGHT:      We introduce BlaBla, an open-source Python library for extracting linguistic features with proven clinical relevance to neurological and psychiatric diseases across many languages.

47, TITLE:      Applying the Transformer to Character-level Transduction
http://arxiv.org/abs/2005.10213
AUTHORS:        Shijie Wu ; Ryan Cotterell ; Mans Hulden
HIGHLIGHT:      The model offers other benefits as well: It trains faster and has fewer parameters.

48, TITLE:      Intra- and Inter-Action Understanding via Temporal Action Parsing
http://arxiv.org/abs/2005.10229
AUTHORS:        Dian Shao ; Yue Zhao ; Bo Dai ; Dahua Lin
COMMENTS:       CVPR 2020 Poster; Project page: https://sdolivia.github.io/TAPOS/
HIGHLIGHT:      Our study shows that a sport activity usually consists of multiple sub-actions and that the awareness of such temporal structures is beneficial to action recognition.

49, TITLE:      Compute-Bound and Low-Bandwidth Distributed 3D Graph-SLAM
http://arxiv.org/abs/2005.10222
AUTHORS:        Jincheng Zhang ; Andrew R. Willis ; Jamie Godwin
HIGHLIGHT:      This article describes a new approach for distributed 3D SLAM map building.

50, TITLE:      Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere
http://arxiv.org/abs/2005.10242
AUTHORS:        Tongzhou Wang ; Phillip Isola
HIGHLIGHT:      In this work, we identify two key properties related to the contrastive loss: (1) alignment (closeness) of features from positive pairs, and (2) uniformity of the induced distribution of the (normalized) features on the hypersphere.

51, TITLE:      Sentence level estimation of psycholinguistic norms using joint multidimensional annotations
http://arxiv.org/abs/2005.10232
AUTHORS:        Anil Ramakrishna ; Shrikanth Narayanan
COMMENTS:       5 pages, 4 figures, submitted to Interspeech 2020
HIGHLIGHT:      In this work, we present a novel approach to estimate the psycholinguistic norms at sentence level.

52, TITLE:      What makes for good views for contrastive learning
http://arxiv.org/abs/2005.10243
AUTHORS:        Yonglong Tian ; Chen Sun ; Ben Poole ; Dilip Krishnan ; Cordelia Schmid ; Phillip Isola
COMMENTS:       submitted to ECCV 2020
HIGHLIGHT:      In this paper, we use empirical analysis to better understand the importance of view selection, and argue that we should reduce the mutual information (MI) between views while keeping task-relevant information intact.

53, TITLE:      Deep learning with 4D spatio-temporal data representations for OCT-based force estimation
http://arxiv.org/abs/2005.10033
AUTHORS:        Nils Gessert ; Marcel Bengs ; Matthias Schlüter ; Alexander Schlaefer
COMMENTS:       Accepted for publication in Medical Image Analysis
HIGHLIGHT:      In this work, we extend the problem of deep learning-based force estimation to 4D spatio-temporal data with streams of 3D OCT volumes.

54, TITLE:      Leveraging Graph to Improve Abstractive Multi-Document Summarization
http://arxiv.org/abs/2005.10043
AUTHORS:        Wei Li ; Xinyan Xiao ; Jiachen Liu ; Hua Wu ; Haifeng Wang ; Junping Du
COMMENTS:       Accepted by ACL2020
HIGHLIGHT:      In this paper, we develop a neural abstractive multi-document summarization (MDS) model which can leverage well-known graph representations of documents such as similarity graph and discourse graph, to more effectively process multiple input documents and produce abstractive summaries.

55, TITLE:      Data Consistent CT Reconstruction from Insufficient Data with Learned Prior Images
http://arxiv.org/abs/2005.10034
AUTHORS:        Yixing Huang ; Alexander Preuhs ; Michael Manhart ; Guenter Lauritsch ; Andreas Maier
COMMENTS:       10 pages, 9 figures
HIGHLIGHT:      In this work, we investigate the robustness of deep learning in CT image reconstruction by showing false negative and false positive lesion cases.

56, TITLE:      Risk of Training Diagnostic Algorithms on Data with Demographic Bias
http://arxiv.org/abs/2005.10050
AUTHORS:        Samaneh Abbasi-Sureshjani ; Ralf Raumanns ; Britt E. J. Michels ; Gerard Schouten ; Veronika Cheplygina
HIGHLIGHT:      In this work, we conduct a survey of the MICCAI 2018 proceedings to investigate the common practice in medical image analysis applications.

57, TITLE:      Map Generation from Large Scale Incomplete and Inaccurate Data Labels
http://arxiv.org/abs/2005.10053
AUTHORS:        Rui Zhang ; Conrad Albrecht ; Wei Zhang ; Xiaodong Cui ; Ulrich Finkler ; David Kung ; Siyuan Lu
COMMENTS:       This paper is accepted by KDD 2020
HIGHLIGHT:      In this paper we present progress in developing an algorithmic pipeline and distributed compute system that automates the process of map creation using high resolution aerial images.

58, TITLE:      Lung Segmentation from Chest X-rays using Variational Data Imputation
http://arxiv.org/abs/2005.10052
AUTHORS:        Raghavendra Selvan ; Erik B. Dam ; Sofus Rischel ; Kaining Sheng ; Mads Nielsen ; Akshay Pai
COMMENTS:       Source code, training data and the trained models are available here: https://github.com/raghavian/lungVAE/
HIGHLIGHT:      In this work, we focus on segmenting lungs from such abnormal CXRs as part of a pipeline aimed at automated risk scoring of COVID-19 from CXRs.

59, TITLE:      Enhancing Word Embeddings with Knowledge Extracted from Lexical Resources
http://arxiv.org/abs/2005.10048
AUTHORS:        Magdalena Biesialska ; Bardia Rafieian ; Marta R. Costa-jussà
COMMENTS:       Accepted to ACL 2020 SRW
HIGHLIGHT:      In this work, we present an effective method for semantic specialization of word vector representations.

60, TITLE:      Early Stage LM Integration Using Local and Global Log-Linear Combination
http://arxiv.org/abs/2005.10049
AUTHORS:        Wilfried Michel ; Ralf Schlüter ; Hermann Ney
COMMENTS:       Submitted to Interspeech 2020
HIGHLIGHT:      In this work, we present a novel method for language model integration into implicit-alignment based sequence-to-sequence models.

61, TITLE:      On embedding Lambek calculus into commutative categorial grammars
http://arxiv.org/abs/2005.10058
AUTHORS:        Sergey Slavnov
HIGHLIGHT:      Therefore different solutions have been proposed in order to enrich ACG with noncommutative constructions.

62, TITLE:      A Large-Scale Multi-Document Summarization Dataset from the Wikipedia Current Events Portal
http://arxiv.org/abs/2005.10070
AUTHORS:        Demian Gholipour Ghalandari ; Chris Hokamp ; Nghia The Pham ; John Glover ; Georgiana Ifrim
COMMENTS:       Camera-ready version for ACL 2020
HIGHLIGHT:      This work presents a new dataset for MDS that is large both in the total number of document clusters and in the size of individual clusters. We build this dataset by leveraging the Wikipedia Current Events Portal (WCEP), which provides concise and neutral human-written summaries of news events, with links to external source articles.

63, TITLE:      Classifying Suspicious Content in Tor Darknet
http://arxiv.org/abs/2005.10086
AUTHORS:        Eduardo Fidalgo Fernandez ; Roberto Andrés Vasco Carofilis ; Francisco Jáñez Martino ; Pablo Blanco Medina
COMMENTS:       To be published on the JNIC 2020 Conference
HIGHLIGHT:      To solve this problem, in this paper, we explore the automatic classification Tor Darknet images using Semantic Attention Keypoint Filtering, a strategy that filters non-significant features at a pixel level that do not belong to the object of interest, by combining saliency maps with Bag of Visual Words (BoVW).

64, TITLE:      Classification of Industrial Control Systems screenshots using Transfer Learning
http://arxiv.org/abs/2005.10098
AUTHORS:        Pablo Blanco Medina ; Eduardo Fidalgo Fernandez ; Enrique Alegre ; Francisco Jáñez Martino ; Roberto A. Vasco-Carofilis ; Víctor Fidalgo Villar
COMMENTS:       To be published on the JNIC 2020 Conference
HIGHLIGHT:      In order to solve this problem, we use transfer learning with five CNN architectures, pre-trained on Imagenet, to determine which one best classifies screenshots obtained from Industrial Controls Systems.

65, TITLE:      Label Efficient Visual Abstractions for Autonomous Driving
http://arxiv.org/abs/2005.10091
AUTHORS:        Aseem Behl ; Kashyap Chitta ; Aditya Prakash ; Eshed Ohn-Bar ; Andreas Geiger
COMMENTS:       First two authors contributed equally, listed in alphabetical order
HIGHLIGHT:      In this work, we seek to quantify the impact of reducing segmentation annotation costs on learned behavior cloning agents.

66, TITLE:      Perceptual Hashing applied to Tor domains recognition
http://arxiv.org/abs/2005.10090
AUTHORS:        Rubel Biswas ; Roberto A. Vasco-Carofilis ; Eduardo Fidalgo Fernandez ; Francisco Jáñez Martino ; Pablo Blanco Medina
COMMENTS:       To be published on the JNIC 2020 Conference
HIGHLIGHT:      To support this task, we introduce Frequency-Dominant Neighborhood Structure (F-DNS), a new perceptual hashing method for automatically classifying domains by their screenshots.

67, TITLE:      Investigation of Large-Margin Softmax in Neural Language Modeling
http://arxiv.org/abs/2005.10089
AUTHORS:        Jingjing Huo ; Yingbo Gao ; Weiyue Wang ; Ralf Schlüter ; Hermann Ney
COMMENTS:       submitted to INTERSPEECH2020
HIGHLIGHT:      In this work, we are curious to see if introducing large-margins to neural language models would improve the perplexity and consequently word error rate in automatic speech recognition.

68, TITLE:      Fair Outlier Detection
http://arxiv.org/abs/2005.09900
AUTHORS:        Deepak P ; Savitha Sam Abraham
HIGHLIGHT:      In this work, we consider the task of fair outlier detection over multiple multi-valued sensitive attributes (e.g., gender, race, religion, nationality, marital status etc.).



==========Updates to Previous Papers==========
1, TITLE:       Saving the Sonorine: Audio Recovery Using Image Processing and Computer Vision
http://arxiv.org/abs/2005.08944
AUTHORS:        Kai Ji ;  Feng ; Adam Finkelstein
COMMENTS:       Removing a co-author. The co-author did not contribute to the preparation of the manuscript, only background information and advice
HIGHLIGHT:      This paper presents a novel technique to recover audio from sonorines, an early 20th century form of analogue sound storage.

2, TITLE:       Conversational Transfer Learning for Emotion Recognition
http://arxiv.org/abs/1910.04980
AUTHORS:        Devamanyu Hazarika ; Soujanya Poria ; Roger Zimmermann ; Rada Mihalcea
COMMENTS:       Information Fusion
HIGHLIGHT:      We propose an approach, TL-ERC, where we pre-train a hierarchical dialogue model on multi-turn conversations (source) and then transfer its parameters to a conversational emotion classifier (target).

3, TITLE:       Abstract categorial grammars with island constraints and effective decidability
http://arxiv.org/abs/1907.06950
AUTHORS:        Sergey Slavnov
COMMENTS:       This was a premature attempt, sorry
HIGHLIGHT:      We adapt this approach to abstract categorial grammars (ACG).

4, TITLE:       Adapting JPEG XS gains and priorities to tasks and contents
http://arxiv.org/abs/2005.08768
AUTHORS:        Benoit Brummer ; Christophe De Vleeschouwer
COMMENTS:       CLIC at CVPR 2020
HIGHLIGHT:      In this work we show that JPEG XS compression can be adapted to a specific given task and content, such as preserving visual quality on desktop content or maintaining high accuracy in neural network segmentation tasks, by optimizing its gain and priority parameters using the covariance matrix adaptation evolution strategy.

5, TITLE:       Neural Architecture Search for Gliomas Segmentation on Multimodal Magnetic Resonance Imaging
http://arxiv.org/abs/2005.06338
AUTHORS:        Feifan Wang
HIGHLIGHT:      In this work, we propose a neural architecture search (NAS) based solution to brain tumor segmentation tasks on multimodal volumetric MRI scans.

6, TITLE:       Volumetric parcellation of the right ventricle for regional geometric and functional assessment
http://arxiv.org/abs/2003.08423
AUTHORS:        Gabriel Bernardino ; Amir Hodzic ; Helene Langet ; Damien LeGallois ; Mathieu De Craene ; Miguel Angel González Ballester ; Eric Saloux ; Bart Bijnens
HIGHLIGHT:      We present a full assessment of the method's validity and reproducibility.

7, TITLE:       RISE Video Dataset: Recognizing Industrial Smoke Emissions
http://arxiv.org/abs/2005.06111
AUTHORS:        Yen-Chia Hsu ; Ting-Hao 'Kenneth' Huang ; Ting-Yao Hu ; Paul Dille ; Sean Prendi ; Ryan Hoffman ; Anastasia Tsuhlares ; Randy Sargent ; Illah Nourbakhsh
COMMENTS:       Technical report
HIGHLIGHT:      We introduce RISE, the first large-scale video dataset for Recognizing Industrial Smoke Emissions.

8, TITLE:       Unlocking New York City Crime Insights using Relational Database Embeddings
http://arxiv.org/abs/2005.09617
AUTHORS:        Apoorva Nitsure ; Rajesh Bordawekar ; Jose Neves
HIGHLIGHT:      This paper demonstrates the use of the AI-Powered Database (AI-DB) in identifying non-obvious patterns in crime data that could serve as an aid to predictive policing measures.

9, TITLE:       Transferable Cost-Aware Security Policy Implementation for Malware Detection Using Deep Reinforcement Learning
http://arxiv.org/abs/1905.10517
AUTHORS:        Yoni Birman ; Shaked Hindi ; Gilad Katz ; Asaf Shabtai
HIGHLIGHT:      In this study, we propose SPIREL, a reinforcement learning-based method for cost-effective malware detection.

10, TITLE:      Efficient adjustment sets in causal graphical models with hidden variables
http://arxiv.org/abs/2004.10521
AUTHORS:        Ezequiel Smucler ; Facundo Sapienza ; Andrea Rotnitzky
COMMENTS:       Fixed typo in the definition of the preorder in page 7 (Z and Z' were backwards)
HIGHLIGHT:      We provide polynomial time algorithms to compute the globally optimal (when it exists), optimal minimal, and optimal minimum adjustment sets.

11, TITLE:      Security of Facial Forensics Models Against Adversarial Attacks
http://arxiv.org/abs/1911.00660
AUTHORS:        Rong Huang ; Fuming Fang ; Huy H. Nguyen ; Junichi Yamagishi ; Isao Echizen
COMMENTS:       Accepted by ICIP 2020
HIGHLIGHT:      We investigated several DNN-based forgery forensics models (FFMs) to examine whether they are secure against adversarial attacks.

12, TITLE:      How Good is Artificial Intelligence at Automatically Answering Consumer Questions Related to Alzheimer's Disease?
http://arxiv.org/abs/1908.10678
AUTHORS:        Krishna B. Soundararajan ; Sunyang Fu ; Luke A. Carlson ; Rebecca A. Smith ; David S. Knopman ; Hongfang Liu ; Yanshan Wang
HIGHLIGHT:      Due to recent advancement in Artificial Intelligence (AI), particularly Natural Language Processing (NLP), we propose to utilize AI to automatically generate answers to AD-related consumer questions posted by caregivers and evaluate how good AI is at answering those questions.

13, TITLE:      SPECTER: Document-level Representation Learning using Citation-informed Transformers
http://arxiv.org/abs/2004.07180
AUTHORS:        Arman Cohan ; Sergey Feldman ; Iz Beltagy ; Doug Downey ; Daniel S. Weld
COMMENTS:       ACL 2020
HIGHLIGHT:      We propose SPECTER, a new method to generate document-level embedding of scientific documents based on pretraining a Transformer language model on a powerful signal of document-level relatedness: the citation graph.

14, TITLE:      Experience Augmentation: Boosting and Accelerating Off-Policy Multi-Agent Reinforcement Learning
http://arxiv.org/abs/2005.09453
AUTHORS:        Zhenhui Ye ; Yining Chen ; Guanghua Song ; Bowei Yang ; Shen Fan
COMMENTS:       10 pages, 4 figures, 4 tables
HIGHLIGHT:      We present a novel technique called Experience Augmentation, which enables a time-efficient and boosted learning based on a fast, fair and thorough exploration to the environment.

15, TITLE:      Learning to Blindly Assess Image Quality in the Laboratory and Wild
http://arxiv.org/abs/1907.00516
AUTHORS:        Weixia Zhang ; Kede Ma ; Guangtao Zhai ; Xiaokang Yang
COMMENTS:       Accepted by ICIP2020
HIGHLIGHT:      To face the cross-distortion-scenario challenge, we develop a BIQA model and an approach of training it on multiple IQA databases (of different distortion scenarios) simultaneously.

16, TITLE:      A Cross-Modal Image Fusion Theory Guided by Human Visual Characteristics
http://arxiv.org/abs/1912.08577
AUTHORS:        Aiqing Fang ; Xinbo Zhao ; Yanning Zhang
HIGHLIGHT:      Inspired by the characteristics of human visual perception, we propose a robust multi-task auxiliary learning optimization image fusion theory.

17, TITLE:      4D Semantic Cardiac Magnetic Resonance Image Synthesis on XCAT Anatomical Model
http://arxiv.org/abs/2002.07089
AUTHORS:        Samaneh Abbasi-Sureshjani ; Sina Amirrajab ; Cristian Lorenz ; Juergen Weese ; Josien Pluim ; Marcel Breeuwer
COMMENTS:       Accepted to MIDL 2020
HIGHLIGHT:      We propose a hybrid controllable image generation method to synthesize anatomically meaningful 3D+t labeled Cardiac Magnetic Resonance (CMR) images.

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

19, TITLE:      Fine-grained Financial Opinion Mining: A Survey and Research Agenda
http://arxiv.org/abs/2005.01897
AUTHORS:        Chung-Chi Chen ; Hen-Hsen Huang ; Hsin-Hsi Chen
HIGHLIGHT:      Fine-grained Financial Opinion Mining: A Survey and Research Agenda

20, TITLE:      Old is Gold: Redefining the Adversarially Learned One-Class Classifier Training Paradigm
http://arxiv.org/abs/2004.07657
AUTHORS:        Muhammad Zaigham Zaheer ; Jin-ha Lee ; Marcella Astrid ; Seung-Ik Lee
COMMENTS:       Accepted at the Conference on Computer Vision and Pattern Recognition CVPR 2020
HIGHLIGHT:      In this study, we propose a framework that effectively generates stable results across a wide range of training steps and allows us to use both the generator and the discriminator of an adversarial model for efficient and robust anomaly detection.

21, TITLE:      On the Unusual Effectiveness of Type-aware Mutations for Testing SMT Solvers
http://arxiv.org/abs/2004.08799
AUTHORS:        Dominik Winterer ; Chengyu Zhang ; Zhendong Su
HIGHLIGHT:      We propose type-aware operator mutation, a simple, but unusually effective approach for testing SMT solvers.

22, TITLE:      Finding Universal Grammatical Relations in Multilingual BERT
http://arxiv.org/abs/2005.04511
AUTHORS:        Ethan A. Chi ; John Hewitt ; Christopher D. Manning
COMMENTS:       To appear in ACL 2020; Farsi typo corrected
HIGHLIGHT:      Motivated by these results, we present an unsupervised analysis method that provides evidence mBERT learns representations of syntactic dependency labels, in the form of clusters which largely agree with the Universal Dependencies taxonomy.

23, TITLE:      On the stable recovery of deep structured linear networks under sparsity constraints
http://arxiv.org/abs/1706.00342
AUTHORS:        Francois Malgouyres
HIGHLIGHT:      As an illustration, we detail the analysis and provide sharp stability guarantees for convolutional linear network under sparsity prior.

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

25, TITLE:      CODA-19: Reliably Annotating Research Aspects on 10,000+ CORD-19 Abstracts Using a Non-Expert Crowd
http://arxiv.org/abs/2005.02367
AUTHORS:        Ting-Hao 'Kenneth' Huang ; Chieh-Yang Huang ; Chien-Kuang Cornelia Ding ; Yen-Chia Hsu ; C. Lee Giles
COMMENTS:       CODA-19: COVID-19 Research Aspect Dataset: https://github.com/windx0303/CODA-19
HIGHLIGHT:      This paper introduces CODA-19, a human-annotated dataset that codes the Background, Purpose, Method, Finding/Contribution, and Other sections of 10,966 English abstracts in the COVID-19 Open Research Dataset.

26, TITLE:      SKEP: Sentiment Knowledge Enhanced Pre-training for Sentiment Analysis
http://arxiv.org/abs/2005.05635
AUTHORS:        Hao Tian ; Can Gao ; Xinyan Xiao ; Hao Liu ; Bolei He ; Hua Wu ; Haifeng Wang ; Feng Wu
COMMENTS:       Accepted by ACL2020
HIGHLIGHT:      In this paper, we introduce Sentiment Knowledge Enhanced Pre-training (SKEP) in order to learn a unified sentiment representation for multiple sentiment analysis tasks.

27, TITLE:      Portrait Shadow Manipulation
http://arxiv.org/abs/2005.08925
AUTHORS:        Xuaner Cecilia Zhang ; Jonathan T. Barron ; Yun-Ta Tsai ; Rohit Pandey ; Xiuming Zhang ; Ren Ng ; David E. Jacobs
COMMENTS:       (updated version); SIGGRAPH 2020;Project webpage: https://people.eecs.berkeley.edu/~cecilia77/project-pages/portrait Video: https://youtu.be/M_qYTXhzyac
HIGHLIGHT:      In this paper, we present a computational approach that gives casual photographers some of this control, thereby allowing poorly-lit portraits to be relit post-capture in a realistic and easily-controllable way. To train our first network we construct a dataset of real-world portraits wherein synthetic foreign shadows are rendered onto the face, and we show that our network learns to remove those unwanted shadows.

28, TITLE:      Text Classification Algorithms: A Survey
http://arxiv.org/abs/1904.08067
AUTHORS:        Kamran Kowsari ; Kiana Jafari Meimandi ; Mojtaba Heidarysafa ; Sanjana Mendu ; Laura E. Barnes ; Donald E. Brown
HIGHLIGHT:      In this paper, a brief overview of text classification algorithms is discussed.

29, TITLE:      Vid2Curve: Simultaneous Camera Motion Estimation and Thin Structure Reconstruction from an RGB Video
http://arxiv.org/abs/2005.03372
AUTHORS:        Peng Wang ; Lingjie Liu ; Nenglun Chen ; Hung-Kuo Chu ; Christian Theobalt ; Wenping Wang
COMMENTS:       Accepted by SIGGRAPH 2020
HIGHLIGHT:      Specifically, we present a new curve-based approach to estimate accurate camera poses by establishing correspondences between featureless thin objects in the foreground in consecutive video frames, without requiring visual texture in the background scene to lock on.

30, TITLE:      The Communication Complexity of Set Intersection and Multiple Equality Testing
http://arxiv.org/abs/1908.11825
AUTHORS:        Dawei Huang ; Seth Pettie ; Yixiang Zhang ; Zhijun Zhang
COMMENTS:       44 pages
HIGHLIGHT:      In this paper we explore fundamental problems in randomized communication complexity such as computing Set Intersection on sets of size $k$ and Equality Testing between vectors of length $k$.

31, TITLE:      A Convolutional Neural Network-based Patent Image Retrieval Method for Design Ideation
http://arxiv.org/abs/2003.08741
AUTHORS:        Shuo Jiang ; Jianxi Luo ; Guillermo Ruiz Pava ; Jie Hu ; Christopher L. Magee
COMMENTS:       11 pages, 11 figures
HIGHLIGHT:      Herein, we propose a convolutional neural network (CNN)-based patent image retrieval method.

32, TITLE:      Analysis of Railway Accidents' Narratives Using Deep Learning
http://arxiv.org/abs/1810.07382
AUTHORS:        Mojtaba Heidarysafa ; Kamran Kowsari ; Laura E. Barnes ; Donald E. Brown
COMMENTS:       accepted in IEEE International Conference on Machine Learning and Applications (IEEE ICMLA)
HIGHLIGHT:      To address these questions, we applied deep learning methods together with powerful word embeddings such as Word2Vec and GloVe to classify accident cause values for the primary cause field using the text in the narratives.

33, TITLE:      Pika parsing: parsing in reverse solves the left recursion and error recovery problems
http://arxiv.org/abs/2005.06444
AUTHORS:        Luke A. D. Hutchison
COMMENTS:       Submitted to ACM
HIGHLIGHT:      Pika parsing: parsing in reverse solves the left recursion and error recovery problems

34, TITLE:      Computing rank of finite algebraic structures with limited nondeterminism
http://arxiv.org/abs/1406.0879
AUTHORS:        Jeffrey Finkelstein
COMMENTS:       Lemma 2.1 incorrectly claims composition of nondeterministic functions
HIGHLIGHT:      We reduce the best upper bounds on the complexity of computing rank for groups and for quasigroups.

35, TITLE:      Learning To Navigate The Synthetically Accessible Chemical Space Using Reinforcement Learning
http://arxiv.org/abs/2004.12485
AUTHORS:        Sai Krishna Gottipati ; Boris Sattarov ; Sufeng Niu ; Yashaswi Pathak ; Haoran Wei ; Shengchao Liu ; Karam M. J. Thomas ; Simon Blackburn ; Connor W. Coley ; Jian Tang ; Sarath Chandar ; Yoshua Bengio
COMMENTS:       added the statistics of top-100 compounds used logP metric with scaled components added values of the initial reactants to the box plots some values in tables are recalculated due to the inconsistent environments on different machines. corresponding benchmarks were rerun with the requirements on github. no significant changes in the results. corrected figures in the Appendix
HIGHLIGHT:      In this work, we propose a novel forward synthesis framework powered by reinforcement learning (RL) for de novo drug design, Policy Gradient for Forward Synthesis (PGFS), that addresses this challenge by embedding the concept of synthetic accessibility directly into the de novo drug design system.

36, TITLE:      T-Net: Nested encoder-decoder architecture for the main vessel segmentation in coronary angiography
http://arxiv.org/abs/1905.04197
AUTHORS:        Tae Joon Jun ; Jihoon Kweon ; Young-Hak Kim ; Daeyoung Kim
COMMENTS:       Neural Networks, Accepted
HIGHLIGHT:      In this paper, we proposed T-Net containing a small encoder-decoder inside the encoder-decoder structure (EDiED).

37, TITLE:      Ultra Fast Structure-aware Deep Lane Detection
http://arxiv.org/abs/2004.11757
AUTHORS:        Zequn Qin ; Huanyu Wang ; Xi Li
HIGHLIGHT:      Motivated by this observation, we propose a novel, simple, yet effective formulation aiming at extremely fast speed and challenging scenarios.

38, TITLE:      Unsupervised Multimodal Hashing for Cross-modal retrieval
http://arxiv.org/abs/1904.00726
AUTHORS:        Jun Yu ; Xiao-Jun Wu
COMMENTS:       The paper is under consideration at IEEE Transaction on Big Data
HIGHLIGHT:      In this paper, we proposed a novel unsupervised hashing learning method to cope with the situation where massive unlabeled data is obtained easily for the era of big data.

39, TITLE:      An IoT Healthcare Framework for Diagnosis of Breast Cancer using Hybrid Transfer Learning
http://arxiv.org/abs/2003.13503
AUTHORS:        Aditya Khamparia ; Subrato Bharati ; Prajoy Podder ; Le Minh Hieu ; Dang N. H. Thanh ; Ugur Erkan
COMMENTS:       24 pages, 11 figures
HIGHLIGHT:      Modified VGG (MVGG), residual network, mobile network is proposed and implemented in this paper.

40, TITLE:      Applications of Probabilistic Programming (Master's thesis, 2015)
http://arxiv.org/abs/1606.00075
AUTHORS:        Yura N Perov
COMMENTS:       Supervisor: Frank Wood. The thesis was prepared in the Department of Engineering Science at the University of Oxford
HIGHLIGHT:      In Chapter 2 we present an approach to automatic discovery of samplers in the form of probabilistic programs.

41, TITLE:      Localizing Firearm Carriers by Identifying Human-Object Pairs
http://arxiv.org/abs/2005.09329
AUTHORS:        Abdul Basit ; Muhammad Akhtar Munir ; Mohsen Ali ; Arif Mahmood
COMMENTS:       5 pages, accepted in IEEE ICIP 2020
HIGHLIGHT:      We present a novel approach to address this problem, by defining human-object interaction (and non-interaction) bounding boxes.

42, TITLE:      Density-Adaptive Kernel based Efficient Reranking Approaches for Person Reidentification
http://arxiv.org/abs/1805.07698
AUTHORS:        Ruo-Pei Guo ; Chun-Guang Li ; Yonghua Li ; Jiaru Lin ; Jun Guo
COMMENTS:       39 pages, 18 figures and 12 tables. This paper is an extended version of our preliminary work on ICPR 2018
HIGHLIGHT:      In this paper, we propose to exploit a density-adaptive smooth kernel technique to achieve efficient and effective reranking.

43, TITLE:      A Computer-Aided Diagnosis System Using Artificial Intelligence for Hip Fractures -Multi-Institutional Joint Development Research-
http://arxiv.org/abs/2003.12443
AUTHORS:        Yoichi Sato ; Yasuhiko Takegami ; Takamune Asamoto ; Yutaro Ono ; Tsugeno Hidetoshi ; Ryosuke Goto ; Akira Kitamura ; Seiwa Honda
COMMENTS:       9 pages, 4 tables, 7 figures. / author's homepage : https://www.fracture-ai.org
HIGHLIGHT:      [Conclusions] The CAD system using deep learning model which we developed was able to diagnose hip fracture in the plane X-ray with the high accuracy, and it was possible to present the decision reason.

44, TITLE:      Shortcut Learning in Deep Neural Networks
http://arxiv.org/abs/2004.07780
AUTHORS:        Robert Geirhos ; Jörn-Henrik Jacobsen ; Claudio Michaelis ; Richard Zemel ; Wieland Brendel ; Matthias Bethge ; Felix A. Wichmann
COMMENTS:       perspective article
HIGHLIGHT:      In this perspective we seek to distil how many of deep learning's problem can be seen as different symptoms of the same underlying problem: shortcut learning. Based on these observations, we develop a set of recommendations for model interpretation and benchmarking, highlighting recent advances in machine learning to improve robustness and transferability from the lab to real-world applications.

45, TITLE:      Spatial-Temporal Graph Convolutional Networks for Sign Language Recognition
http://arxiv.org/abs/1901.11164
AUTHORS:        Cleison Correia de Amorim ; David Macêdo ; Cleber Zanchettin
HIGHLIGHT:      We propose a new approach of Spatial-Temporal Graph Convolutional Network to sign language recognition based on the human skeletal movements. Additionally, we present a new dataset of human skeletons for sign language based on ASLLVD to contribute to future related studies.

46, TITLE:      From Videos to URLs: A Multi-Browser Guide To Extract User's Behavior with Optical Character Recognition
http://arxiv.org/abs/1811.06193
AUTHORS:        Mojtaba Heidarysafa ; James Reed ; Kamran Kowsari ; April Celeste R. Leviton ; Janet I. Warren ; Donald E. Brown
HIGHLIGHT:      In this paper, we present an image-processing based method to extract domains which are visited by a participant over multiple browsers during a lab session.

47, TITLE:      The AVA-Kinetics Localized Human Actions Video Dataset
http://arxiv.org/abs/2005.00214
AUTHORS:        Ang Li ; Meghana Thotakuri ; David A. Ross ; João Carreira ; Alexander Vostrikov ; Andrew Zisserman
COMMENTS:       8 pages, 8 figures
HIGHLIGHT:      This paper describes the AVA-Kinetics localized human actions video dataset.

48, TITLE:      GRIP++: Enhanced Graph-based Interaction-aware Trajectory Prediction for Autonomous Driving
http://arxiv.org/abs/1907.07792
AUTHORS:        Xin Li ; Xiaowen Ying ; Mooi Choo Chuah
HIGHLIGHT:      Hence, in this paper, we describe an improved scheme called GRIP++ where we use both fixed and dynamic graphs for trajectory predictions of different types of traffic agents.

49, TITLE:      Question-Driven Summarization of Answers to Consumer Health Questions
http://arxiv.org/abs/2005.09067
AUTHORS:        Max Savery ; Asma Ben Abacha ; Soumya Gayen ; Dina Demner-Fushman
HIGHLIGHT:      Using answers provided by the National Library of Medicine's consumer health question answering system, we present the MEDIQA Answer Summarization dataset, the first summarization collection containing question-driven summaries of answers to consumer health questions.

50, TITLE:      A Philosophy of Data
http://arxiv.org/abs/2004.09990
AUTHORS:        Alexander M. Mussgnug
HIGHLIGHT:      We argue that while this discourse on data ethics is of critical importance, it is missing one fundamental point: If more and more efforts in business, government, science, and our daily lives are data-driven, we should pay more attention to what exactly we are driven by.

51, 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, 5 figures, submitted to IEEE Robotics and Automation Letters, revision including additional real-world experiments and improved visualization
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.

52, TITLE:      Dual-Sampling Attention Network for Diagnosis of COVID-19 from Community Acquired Pneumonia
http://arxiv.org/abs/2005.02690
AUTHORS:        Xi Ouyang ; Jiayu Huo ; Liming Xia ; Fei Shan ; Jun Liu ; Zhanhao Mo ; Fuhua Yan ; Zhongxiang Ding ; Qi Yang ; Bin Song ; Feng Shi ; Huan Yuan ; Ying Wei ; Xiaohuan Cao ; Yaozong Gao ; Dijia Wu ; Qian Wang ; Dinggang Shen
COMMENTS:       accepted by IEEE Transactions on Medical Imaging, 2020
HIGHLIGHT:      In particular, we propose a novel online attention module with a 3D convolutional network (CNN) to focus on the infection regions in lungs when making decisions of diagnoses.

53, TITLE:      Deep Manifold Embedding for Hyperspectral Image Classification
http://arxiv.org/abs/1912.11264
AUTHORS:        Zhiqiang Gong ; Weidong Hu ; Xiaoyong Du ; Ping Zhong ; Panhe Hu
COMMENTS:       Submitted to ISPRS
HIGHLIGHT:      To tackle this problem, this work develops a novel deep manifold embedding method(DMEM) for hyperspectral image classification.

54, TITLE:      BLEURT: Learning Robust Metrics for Text Generation
http://arxiv.org/abs/2004.04696
AUTHORS:        Thibault Sellam ; Dipanjan Das ; Ankur P. Parikh
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      We propose BLEURT, a learned evaluation metric based on BERT that can model human judgments with a few thousand possibly biased training examples.

55, TITLE:      DeepFaceLab: A simple, flexible and extensible face swapping framework
http://arxiv.org/abs/2005.05535
AUTHORS:        Ivan Perov ; Daiheng Gao ; Nikolay Chervoniy ; Kunlin Liu ; Sugasa Marangonda ; Chris Umé ; Mr. Dpfks ; Carl Shift Facenheim ; Luis RP ; Jian Jiang ; Sheng Zhang ; Pingyu Wu ; Bo Zhou ; Weiming Zhang
HIGHLIGHT:      In this paper, we detail the principles that drive the implementation of DeepFaceLab and introduce the pipeline of it, through which every aspect of the pipeline can be modified painlessly by users to achieve their customization purpose, and it's noteworthy that DeepFaceLab could achieve results with high fidelity and indeed indiscernible by mainstream forgery detection approaches.

56, TITLE:      Rigid Matrices From Rectangular PCPs
http://arxiv.org/abs/2005.03123
AUTHORS:        Amey Bhangale ; Prahladh Harsha ; Orr Paradise ; Avishay Tal
COMMENTS:       36 pages, 3 figures
HIGHLIGHT:      We introduce a variant of PCPs, that we refer to as rectangular PCPs, wherein proofs are thought of as square matrices, and the random coins used by the verifier can be partitioned into two disjoint sets, one determining the row of each query and the other determining the *column*.