2020.04.07.txt

==========New Papers==========
1, TITLE:       On Sharpness of Error Bounds for Multivariate Neural Network Approximation
http://arxiv.org/abs/2004.02203
AUTHORS:        Steffen Goebbels
COMMENTS:       ongoing work
HIGHLIGHT:      On Sharpness of Error Bounds for Multivariate Neural Network Approximation

2, TITLE:       The two-echelon routing problem with truck and drones
http://arxiv.org/abs/2004.02275
AUTHORS:        Minh Hoàng Hà ; Lam Vu ; Duy Manh Vu
COMMENTS:       29 pages, 4 figures
HIGHLIGHT:      In this paper, we study novel variants of the well-known two-echelon vehicle routing problem in which a truck works on the first echelon to transport parcels and a fleet of drones to intermediate depots while in the second echelon, the drones are used to deliver parcels from intermediate depots to customers.

3, TITLE:       Generating Rationales in Visual Question Answering
http://arxiv.org/abs/2004.02032
AUTHORS:        Hammad A. Ayyubi ; Md. Mehrab Tanjim ; Julian J. McAuley ; Garrison W. Cottrell
HIGHLIGHT:      Generating Rationales in Visual Question Answering

4, TITLE:       Morphological Computation and Learning to Learn In Natural Intelligent Systems And AI
http://arxiv.org/abs/2004.02304
AUTHORS:        Gordana Dodig-Crnkovic
COMMENTS:       5 pages, no figures, AISB 2020 conference
HIGHLIGHT:      Learning from nature is a two-way process as discussed in [2][3][4], computing is learning from neuroscience, while neuroscience is quickly adopting information processing models.

5, TITLE:       Verifying Recurrent Neural Networks using Invariant Inference
http://arxiv.org/abs/2004.02462
AUTHORS:        Yuval Jacoby ; Clark Barrett ; Guy Katz
HIGHLIGHT:      Here, we propose a novel approach for verifying properties of a widespread variant of neural networks, called recurrent neural networks.

6, TITLE:       A Dependency Syntactic Knowledge Augmented Interactive Architecture for End-to-End Aspect-based Sentiment Analysis
http://arxiv.org/abs/2004.01951
AUTHORS:        Yunlong Liang ; Fandong Meng ; Jinchao Zhang ; Jinan Xu ; Yufeng Chen ; Jie Zhou
COMMENTS:       Code: https://github.com/XL2248/DREGCN
HIGHLIGHT:      In this paper, we thus propose a novel dependency syntactic knowledge augmented interactive architecture with multi-task learning for end-to-end ABSA.

7, TITLE:       An Iterative Knowledge Transfer Network with Routing for Aspect-based Sentiment Analysis
http://arxiv.org/abs/2004.01935
AUTHORS:        Yunlong Liang ; Fandong Meng ; Jinchao Zhang ; Jinan Xu ; Yufeng Chen ; Jie Zhou
COMMENTS:       Code: https://github.com/XL2248/IKTN
HIGHLIGHT:      To address these issues, we propose a novel Iterative Knowledge Transfer Network (IKTN) for the end-to-end ABSA.

8, TITLE:       Pre-Trained and Attention-Based Neural Networks for Building Noetic Task-Oriented Dialogue Systems
http://arxiv.org/abs/2004.01940
AUTHORS:        Jia-Chen Gu ; Tianda Li ; Quan Liu ; Xiaodan Zhu ; Zhen-Hua Ling ; Yu-Ping Ruan
COMMENTS:       Accepted by AAAI 2020, Workshop on DSTC8
HIGHLIGHT:      This paper describes our systems that are evaluated on all subtasks under this challenge.

9, TITLE:       Prerequisites for Explainable Machine Reading Comprehension: A Position Paper
http://arxiv.org/abs/2004.01912
AUTHORS:        Saku Sugawara ; Pontus Stenetorp ; Akiko Aizawa
HIGHLIGHT:      To this end, this position paper provides a theoretical basis for the design of MRC based on psychology and psychometrics and summarizes it in terms of the requirements for explainable MRC.

10, TITLE:      BAE: BERT-based Adversarial Examples for Text Classification
http://arxiv.org/abs/2004.01970
AUTHORS:        Siddhant Garg ; Goutham Ramakrishnan
HIGHLIGHT:      We present BAE, a powerful black box attack for generating grammatically correct and semantically coherent adversarial examples.

11, TITLE:      "None of the Above":Measure Uncertainty in Dialog Response Retrieval
http://arxiv.org/abs/2004.01926
AUTHORS:        Yulan Feng ; Shikib Mehri ; Maxine Eskenazi ; Tiancheng Zhao
HIGHLIGHT:      This paper discusses the importance of uncovering uncertainty in end-to-end dialog tasks, and presents our experimental results on uncertainty classification on the Ubuntu Dialog Corpus.

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

13, TITLE:      Image-based phenotyping of diverse Rice (Oryza Sativa L.) Genotypes
http://arxiv.org/abs/2004.02498
AUTHORS:        Mukesh Kumar Vishal ; Dipesh Tamboli ; Abhijeet Patil ; Rohit Saluja ; Biplab Banerjee ; Amit Sethi ; Dhandapani Raju ; Sudhir Kumar ; R N Sahoo ; Viswanathan Chinnusamy ; J Adinarayana
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      We trained You Only Look Once (YOLO) deep learning algorithm for leaves tips detection and to estimate the number of leaves in a rice plant.

14, TITLE:      Exploration of Input Patterns for Enhancing the Performance of Liquid State Machines
http://arxiv.org/abs/2004.02540
AUTHORS:        Shasha Guo ; Lianhua Qu ; Lei Wang ; Xulong Tang ; Shuo Tian ; Shiming Li ; Weixia Xu
HIGHLIGHT:      We explore the influence of input scale reduction on LSM instead.

15, TITLE:      Vanishing Point Guided Natural Image Stitching
http://arxiv.org/abs/2004.02478
AUTHORS:        Kai Chen ; Jian Yao ; Jingmin Tu ; Yahui Liu ; Yinxuan Li ; Li Li
COMMENTS:       13 pages, 16 figures
HIGHLIGHT:      In this paper, we propose a novel natural image stitching method, which takes into account the guidance of vanishing points to tackle the mentioned failures.

16, TITLE:      On-device Filtering of Social Media Images for Efficient Storage
http://arxiv.org/abs/2004.02489
AUTHORS:        Dhruval Jain ; DP Mohanty ; Sanjeev Roy ; Naresh Purre ; Sukumar Moharana
HIGHLIGHT:      To address this, we propose a novel method based on Convolutional Neural Networks (CNNs) for the on-device filtering of social media images by classifying these synthetic images and allowing the user to delete them in one go.

17, TITLE:      Cascaded Deep Video Deblurring Using Temporal Sharpness Prior
http://arxiv.org/abs/2004.02501
AUTHORS:        Jinshan Pan ; Haoran Bai ; Jinhui Tang
COMMENTS:       CVPR 2020. The code is available at https://github.com/csbhr/CDVD-TSP
HIGHLIGHT:      We present a simple and effective deep convolutional neural network (CNN) model for video deblurring.

18, TITLE:      GANSpace: Discovering Interpretable GAN Controls
http://arxiv.org/abs/2004.02546
AUTHORS:        Erik Härkönen ; Aaron Hertzmann ; Jaakko Lehtinen ; Sylvain Paris
HIGHLIGHT:      This paper describes a simple technique to analyze Generative Adversarial Networks (GANs) and create interpretable controls for image synthesis, such as change of viewpoint, aging, lighting, and time of day.

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

20, TITLE:      Any-Shot Sequential Anomaly Detection in Surveillance Videos
http://arxiv.org/abs/2004.02072
AUTHORS:        Keval Doshi ; Yasin Yilmaz
COMMENTS:       Accepted to CVPR 2020: Workshop on Continual Learning in Computer Vision
HIGHLIGHT:      Motivated by these research gaps, we propose an online anomaly detection method for surveillance videos using transfer learning and any-shot learning, which in turn significantly reduces the training complexity and provides a mechanism that can detect anomalies using only a few labeled nominal examples.

21, TITLE:      gDLS*: Generalized Pose-and-Scale Estimation Given Scale and Gravity Priors
http://arxiv.org/abs/2004.02052
AUTHORS:        Victor Fragoso ; Joseph DeGol ; Gang Hua
HIGHLIGHT:      We present gDLS*, a generalized-camera-model pose-and-scale estimator that utilizes rotation and scale priors.

22, TITLE:      TensorFI: A Flexible Fault Injection Framework for TensorFlow Applications
http://arxiv.org/abs/2004.01743
AUTHORS:        Zitao Chen ; Niranjhana Narayanan ; Bo Fang ; Guanpeng Li ; Karthik Pattabiraman ; Nathan DeBardeleben
COMMENTS:       A preliminary version of this work was published in a workshop
HIGHLIGHT:      In this work, we present TensorFI, a high-level fault injection (FI) framework for TensorFlow-based applications.

23, TITLE:      Theoretical Insights into the Use of Structural Similarity Index In Generative Models and Inferential Autoencoders
http://arxiv.org/abs/2004.01864
AUTHORS:        Benyamin Ghojogh ; Fakhri Karray ; Mark Crowley
COMMENTS:       Accepted (to appear) in International Conference on Image Analysis and Recognition (ICIAR) 2020, Springer
HIGHLIGHT:      Generative models and inferential autoencoders mostly make use of $\ell_2$ norm in their optimization objectives.

24, TITLE:      Unsupervised Domain Adaptation with Progressive Domain Augmentation
http://arxiv.org/abs/2004.01735
AUTHORS:        Kevin Hua ; Yuhong Guo
HIGHLIGHT:      In the paper, we propose a novel unsupervised domain adaptation method based on progressive domain augmentation.

25, TITLE:      On the Convergence Analysis of Asynchronous SGD for Solving Consistent Linear Systems
http://arxiv.org/abs/2004.02163
AUTHORS:        Atal Narayan Sahu ; Aritra Dutta ; Aashutosh Tiwari ; Peter Richtárik
HIGHLIGHT:      In this paper, we propose and analyze a {\it distributed, asynchronous parallel} SGD in light of solving an arbitrary consistent linear system by reformulating the system into a stochastic optimization problem as studied by Richt\'{a}rik and Tak\'{a}\~{c} in [35].

26, TITLE:      PONE: A Novel Automatic Evaluation Metric for Open-Domain Generative Dialogue Systems
http://arxiv.org/abs/2004.02399
AUTHORS:        Tian Lan ; Xian-Ling Mao ; Wei Wei ; Xiaoyan Gao ; Heyan Huang
HIGHLIGHT:      In this paper, we will first measure systematically all kinds of automatic evaluation metrics over the same experimental setting to check which kind is best.

27, TITLE:      An analysis of the utility of explicit negative examples to improve the syntactic abilities of neural language models
http://arxiv.org/abs/2004.02451
AUTHORS:        Hiroshi Noji ; Hiroya Takamura
COMMENTS:       To appear at ACL 2020 (long paper)
HIGHLIGHT:      We explore the utilities of explicit negative examples in training neural language models.

28, TITLE:      Distinguish Confusing Law Articles for Legal Judgment Prediction
http://arxiv.org/abs/2004.02557
AUTHORS:        Nuo Xu ; Pinghui Wang ; Long Chen ; Li Pan ; Xiaoyan Wang ; Junzhou Zhao
COMMENTS:       This work has been accepted by the 58th Annual Meeting of the Association for Computational Linguistics (ACL 2020)
HIGHLIGHT:      In this paper, we present an end-to-end model, LADAN, to solve the task of LJP.

29, TITLE:      SelfORE: Self-supervised Relational Feature Learning for Open Relation Extraction
http://arxiv.org/abs/2004.02438
AUTHORS:        Xuming Hu ; Lijie Wen ; Yusong Xu ; Chenwei Zhang ; Philip S. Yu
HIGHLIGHT:      In this work, we proposed a self-supervised framework named SelfORE, which exploits weak, self-supervised signals by leveraging large pretrained language model for adaptive clustering on contextualized relational features, and bootstraps the self-supervised signals by improving contextualized features in relation classification.

30, TITLE:      Dictionary-based Data Augmentation for Cross-Domain Neural Machine Translation
http://arxiv.org/abs/2004.02577
AUTHORS:        Wei Peng ; Chongxuan Huang ; Tianhao Li ; Yun Chen ; Qun Liu
HIGHLIGHT:      This paper proposes a dictionary-based data augmentation (DDA) method for cross-domain NMT.

31, TITLE:      Grayscale Data Construction and Multi-Level Ranking Objective for Dialogue Response Selection
http://arxiv.org/abs/2004.02421
AUTHORS:        Zibo Lin ; Deng Cai ; Yan Wang ; Xiaojiang Liu ; Hai-Tao Zheng ; Shuming Shi
HIGHLIGHT:      We propose to automatically build training data with grayscale labels.

32, TITLE:      Building a Norwegian Lexical Resource for Medical Entity Recognition
http://arxiv.org/abs/2004.02509
AUTHORS:        Ildikó Pilán ; Pål H. Brekke ; Lilja Øvrelid
HIGHLIGHT:      We present a large Norwegian lexical resource of categorized medical terms.

33, TITLE:      On the Tandem Duplication Distance
http://arxiv.org/abs/2004.02338
AUTHORS:        Ferdinando Cicalese ; Nicolò Pilati
HIGHLIGHT:      In this paper, we significantly improve this result and show that the tandem duplication "distance problem" is NP-hard already for the case of strings over an alphabet of size $\leq 5.

34, TITLE:      Evaluating Multimodal Representations on Visual Semantic Textual Similarity
http://arxiv.org/abs/2004.01894
AUTHORS:        Oier Lopez de Lacalle ; Ander Salaberria ; Aitor Soroa ; Gorka Azkune ; Eneko Agirre
COMMENTS:       Accepted in ECAI-2020, 8 pages, 6 tables, 6 figures
HIGHLIGHT:      The long term goal of our research is to devise multimodal representation techniques that improve current inference capabilities.

35, TITLE:      Knowledge Guided Metric Learning for Few-Shot Text Classification
http://arxiv.org/abs/2004.01907
AUTHORS:        Dianbo Sui ; Yubo Chen ; Binjie Mao ; Delai Qiu ; Kang Liu ; Jun Zhao
HIGHLIGHT:      Inspired by human intelligence, we propose to introduce external knowledge into few-shot learning to imitate human knowledge.

36, TITLE:      CG-BERT: Conditional Text Generation with BERT for Generalized Few-shot Intent Detection
http://arxiv.org/abs/2004.01881
AUTHORS:        Congying Xia ; Chenwei Zhang ; Hoang Nguyen ; Jiawei Zhang ; Philip Yu
COMMENTS:       11 pages, 4 figures
HIGHLIGHT:      In this paper, we formulate a more realistic and difficult problem setup for the intent detection task in natural language understanding, namely Generalized Few-Shot Intent Detection (GFSID).

37, TITLE:      STEP: Sequence-to-Sequence Transformer Pre-training for Document Summarization
http://arxiv.org/abs/2004.01853
AUTHORS:        Yanyan Zou ; Xingxing Zhang ; Wei Lu ; Furu Wei ; Ming Zhou
HIGHLIGHT:      We, therefore, propose STEP (as shorthand for Sequence-to-Sequence Transformer Pre-training), which can be trained on large scale unlabeled documents.

38, TITLE:      Conversational Question Reformulation via Sequence-to-Sequence Architectures and Pretrained Language Models
http://arxiv.org/abs/2004.01909
AUTHORS:        Sheng-Chieh Lin ; Jheng-Hong Yang ; Rodrigo Nogueira ; Ming-Feng Tsai ; Chuan-Ju Wang ; Jimmy Lin
HIGHLIGHT:      This paper presents an empirical study of conversational question reformulation (CQR) with sequence-to-sequence architectures and pretrained language models (PLMs).

39, TITLE:      News-Driven Stock Prediction With Attention-Based Noisy Recurrent State Transition
http://arxiv.org/abs/2004.01878
AUTHORS:        Xiao Liu ; Heyan Huang ; Yue Zhang ; Changsen Yuan
COMMENTS:       12 pages
HIGHLIGHT:      We consider direct modeling of underlying stock value movement sequences over time in the news-driven stock movement prediction.

40, TITLE:      AutoToon: Automatic Geometric Warping for Face Cartoon Generation
http://arxiv.org/abs/2004.02377
AUTHORS:        Julia Gong ; Yannick Hold-Geoffroy ; Jingwan Lu
COMMENTS:       Accepted and presented at WACV 2020; to appear in proceedings of 2020 IEEE Winter Conference on Applications of Computer Vision (WACV). Completed during Julia Gong's internship at Adobe Research
HIGHLIGHT:      In this work, we propose AutoToon, the first supervised deep learning method that yields high-quality warps for the warping component of caricatures.

41, TITLE:      Detecting the Saliency of Remote Sensing Images Based on Sparse Representation of Contrast-weighted Atoms
http://arxiv.org/abs/2004.02428
AUTHORS:        Zhou Huang ; Huai-Xin Chen ; Yun-Zhi Yang ; Chang-Yin Wang ; Bi-Yuan Liu
HIGHLIGHT:      In this paper, a novel saliency detection method is proposed by exploring the sparse representation (SR) of, based on learning, contrast-weighted atoms (LCWA).

42, TITLE:      Deep Space-Time Video Upsampling Networks
http://arxiv.org/abs/2004.02432
AUTHORS:        Jaeyeon Kang ; Younghyun Jo ; Seoung Wug Oh ; Peter Vajda ; Seon Joo Kim
HIGHLIGHT:      In this paper, we investigate the problem of jointly upsampling videos both in space and time, which is becoming more important with advances in display systems.

43, TITLE:      Robust 3D Self-portraits in Seconds
http://arxiv.org/abs/2004.02460
AUTHORS:        Zhe Li ; Tao Yu ; Chuanyu Pan ; Zerong Zheng ; Yebin Liu
COMMENTS:       CVPR 2020, oral
HIGHLIGHT:      In this paper, we propose an efficient method for robust 3D self-portraits using a single RGBD camera.

44, TITLE:      Class Anchor Clustering: a Distance-based Loss for Training Open Set Classifiers
http://arxiv.org/abs/2004.02434
AUTHORS:        Dimity Miller ; Niko Sünderhauf ; Michael Milford ; Feras Dayoub
HIGHLIGHT:      To overcome this limitation, we introduce Class Anchor Clustering (CAC) loss.

45, TITLE:      Neural Architecture Search for Lightweight Non-Local Networks
http://arxiv.org/abs/2004.01961
AUTHORS:        Yingwei Li ; Xiaojie Jin ; Jieru Mei ; Xiaochen Lian ; Linjie Yang ; Cihang Xie ; Qihang Yu ; Yuyin Zhou ; Song Bai ; Alan Yuille
COMMENTS:       CVPR 2020. Project page: https://github.com/LiYingwei/AutoNL
HIGHLIGHT:      We propose AutoNL to overcome the above two obstacles.

46, TITLE:      Optical Flow in Dense Foggy Scenes using Semi-Supervised Learning
http://arxiv.org/abs/2004.01905
AUTHORS:        Wending Yan ; Aashish Sharma ; Robby T. Tan
HIGHLIGHT:      To address the problem, we introduce a semi-supervised deep learning technique that employs real fog images without optical flow ground-truths in the training process.

47, TITLE:      Cross-domain Face Presentation Attack Detection via Multi-domain Disentangled Representation Learning
http://arxiv.org/abs/2004.01959
AUTHORS:        Guoqing Wang ; Hu Han ; Shiguang Shan ; Xilin Chen
COMMENTS:       Accepted by CVPR2020
HIGHLIGHT:      In light of this, we propose an efficient disentangled representation learning for cross-domain face PAD.

48, TITLE:      Weakly-Supervised Mesh-Convolutional Hand Reconstruction in the Wild
http://arxiv.org/abs/2004.01946
AUTHORS:        Dominik Kulon ; Riza Alp Güler ; Iasonas Kokkinos ; Michael Bronstein ; Stefanos Zafeiriou
COMMENTS:       Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2020). Additional resources: https://arielai.com/mesh_hands
HIGHLIGHT:      We introduce a simple and effective network architecture for monocular 3D hand pose estimation consisting of an image encoder followed by a mesh convolutional decoder that is trained through a direct 3D hand mesh reconstruction loss.

49, TITLE:      Understanding (Non-)Robust Feature Disentanglement and the Relationship Between Low- and High-Dimensional Adversarial Attacks
http://arxiv.org/abs/2004.01903
AUTHORS:        Zuowen Wang ; Leo Horne
HIGHLIGHT:      Recent work has put forth the hypothesis that adversarial vulnerabilities in neural networks are due to them overusing "non-robust features" inherent in the training data.

50, TITLE:      Variable Shift SDD: A More Succinct Sentential Decision Diagram
http://arxiv.org/abs/2004.02502
AUTHORS:        Kengo Nakamura ; Shuhei Denzumi ; Masaaki Nishino
COMMENTS:       21 pages; Accepted for SEA2020
HIGHLIGHT:      In this paper, we propose a more succinct variant of SDD named Variable Shift SDD (VS-SDD).

51, TITLE:      Applying Cyclical Learning Rate to Neural Machine Translation
http://arxiv.org/abs/2004.02401
AUTHORS:        Choon Meng Lee ; Jianfeng Liu ; Wei Peng
HIGHLIGHT:      Drawing inspiration from the successful application of cyclical learning rate policy for computer vision related convolutional networks and datasets, we explore how cyclical learning rate can be applied to train transformer-based neural networks for neural machine translation.

52, TITLE:      Weighted Fisher Discriminant Analysis in the Input and Feature Spaces
http://arxiv.org/abs/2004.01857
AUTHORS:        Benyamin Ghojogh ; Milad Sikaroudi ; H. R. Tizhoosh ; Fakhri Karray ; Mark Crowley
COMMENTS:       Accepted (to appear) in International Conference on Image Analysis and Recognition (ICIAR) 2020, Springer
HIGHLIGHT:      In this paper, we propose a cosine-weighted FDA as well as an automatically weighted FDA in which weights are found automatically.

53, TITLE:      Attention-Guided Version of 2D UNet for Automatic Brain Tumor Segmentation
http://arxiv.org/abs/2004.02009
AUTHORS:        Mehrdad Noori ; Ali Bahri ; Karim Mohammadi
COMMENTS:       7 pages, 5 figures, 4 tables, Accepted by ICCKE 2019
HIGHLIGHT:      In order to address the mentioned issues, we design a low-parameter network based on 2D UNet in which we employ two techniques.

54, TITLE:      Segmentation for Classification of Screening Pancreatic Neuroendocrine Tumors
http://arxiv.org/abs/2004.02021
AUTHORS:        Zhuotun Zhu ; Yongyi Lu ; Wei Shen ; Elliot K. Fishman ; Alan L. Yuille
HIGHLIGHT:      This work presents comprehensive results to detect in the early stage the pancreatic neuroendocrine tumors (PNETs), a group of endocrine tumors arising in the pancreas, which are the second common type of pancreatic cancer, by checking the abdominal CT scans. To quantitatively analyze our method, we collect and voxelwisely label a new abdominal CT dataset containing $376$ cases with both arterial and venous phases available for each case, in which $228$ cases were diagnosed with PNETs while the remaining $148$ cases are normal, which is currently the largest dataset for PNETs to the best of our knowledge.

55, TITLE:      Empirical Evaluation of PRNU Fingerprint Variation for Mismatched Imaging Pipelines
http://arxiv.org/abs/2004.01929
AUTHORS:        Sharad Joshi ; Pawel Korus ; Nitin Khanna ; Nasir Memon
HIGHLIGHT:      We show that camera fingerprints exhibit non-negligible variations in this setup, which may lead to unexpected degradation of detection statistics in real-world use-cases.

56, TITLE:      Volumetric Attention for 3D Medical Image Segmentation and Detection
http://arxiv.org/abs/2004.01997
AUTHORS:        Xudong Wang ; Shizhong Han ; Yunqiang Chen ; Dashan Gao ; Nuno Vasconcelos
COMMENTS:       Accepted by MICCAI 2019
HIGHLIGHT:      Volumetric Attention for 3D Medical Image Segmentation and Detection

57, TITLE:      Ontologies in CLARIAH: Towards Interoperability in History, Language and Media
http://arxiv.org/abs/2004.02845
AUTHORS:        Albert Meroño-Peñuela ; Victor de Boer ; Marieke van Erp ; Willem Melder ; Rick Mourits ; Auke Rijpma ; Ruben Schalk ; Richard Zijdeman
COMMENTS:       26 pages
HIGHLIGHT:      In this chapter, we describe the ontologies and tools developed and integrated in the Dutch national project CLARIAH to address these issues across datasets from three fundamental domains or "pillars" of the humanities (linguistics, social and economic history, and media studies) that have paradigmatic data representations (textual corpora, structured data, and multimedia).

58, TITLE:      A new approach for generation of generalized basic probability assignment in the evidence theory
http://arxiv.org/abs/2004.02746
AUTHORS:        Dongdong Wu ; Zijing Liu ; Yongchuan Tang
HIGHLIGHT:      In this paper, a new method is proposed to generate generalized basic probability assignment (GBPA) based on the triangular fuzzy number model under the open world assumption.

59, TITLE:      Answering Complex Queries in Knowledge Graphs with Bidirectional Sequence Encoders
http://arxiv.org/abs/2004.02596
AUTHORS:        Bhushan Kotnis ; Carolin Lawrence ; Mathias Niepert
COMMENTS:       7 pages, 2 figures
HIGHLIGHT:      In this work we address the more ambitious challenge of predicting the answers of conjunctive queries with multiple missing entities. We introduce a new dataset for predicting the answer of conjunctive query and conduct experiments that show \textsc{BiQE} significantly outperforming state of the art baselines.

60, TITLE:      On Evaluating the Quality of Rule-Based Classification Systems
http://arxiv.org/abs/2004.02671
AUTHORS:        Nassim Dehouche
COMMENTS:       ICIC Express Letters Volume 11, Number 10, October 2017
HIGHLIGHT:      In this work, we claim that these two indicators may be insufficient, and additional measures of quality may need to be developed.

61, TITLE:      Trust-based Multiagent Consensus or Weightings Aggregation
http://arxiv.org/abs/2004.02490
AUTHORS:        Bruno Yun ; Madalina Croitoru
HIGHLIGHT:      We introduce a framework for reaching a consensus amongst several agents communicating via a trust network on conflicting information about their environment.

62, TITLE:      Natural language processing for word sense disambiguation and information extraction
http://arxiv.org/abs/2004.02256
AUTHORS:        K. R. Chowdhary
COMMENTS:       150 pages, PhD Thesis
HIGHLIGHT:      The thesis presents a new approach for Word Sense Disambiguation using thesaurus.

63, TITLE:      Speaker Recognition using SincNet and X-Vector Fusion
http://arxiv.org/abs/2004.02219
AUTHORS:        Mayank Tripathi ; Divyanshu Singh ; Seba Susan
COMMENTS:       The 19th International Conference on Artificial Intelligence and Soft Computing
HIGHLIGHT:      In this paper, we propose an innovative approach to perform speaker recognition by fusing two recently introduced deep neural networks (DNNs) namely - SincNet and X-Vector.

64, TITLE:      Semantics of the Unwritten
http://arxiv.org/abs/2004.02251
AUTHORS:        He Bai ; Peng Shi ; Jimmy Lin ; Luchen Tan ; Kun Xiong ; Wen Gao ; Jie Liu ; Ming Li
HIGHLIGHT:      In this article, we will study how those implicit "not read" information such as end-of-paragraph (EOP) and end-of-sequence (EOS) affect the quality of text generation.

65, TITLE:      Syntax-driven Iterative Expansion Language Models for Controllable Text Generation
http://arxiv.org/abs/2004.02211
AUTHORS:        Noe Casas ; José A. R. Fonollosa ; Marta R. Costa-jussà
HIGHLIGHT:      We propose a new paradigm for introducing a syntactic inductive bias into neural language modeling and text generation, where the dependency parse tree is used to drive the Transformer model to generate sentences iteratively, starting from a root placeholder and generating the tokens of the different dependency tree branches in parallel, using either word or subword vocabularies.

66, TITLE:      Stylistic Dialogue Generation via Information-Guided Reinforcement Learning Strategy
http://arxiv.org/abs/2004.02202
AUTHORS:        Yixuan Su ; Deng Cai ; Yan Wang ; Simon Baker ; Anna Korhonen ; Nigel Collier ; Xiaojiang Liu
HIGHLIGHT:      To enable better balance between the content quality and the style, we introduce a new training strategy, know as Information-Guided Reinforcement Learning (IG-RL).

67, TITLE:      Prototype-to-Style: Dialogue Generation with Style-Aware Editing on Retrieval Memory
http://arxiv.org/abs/2004.02214
AUTHORS:        Yixuan Su ; Yan Wang ; Simon Baker ; Deng Cai ; Xiaojiang Liu ; Anna Korhonen ; Nigel Collier
HIGHLIGHT:      To effectively train the proposed model, we propose a new style-aware learning objective as well as a de-noising learning strategy.

68, TITLE:      Improved Pretraining for Domain-specific Contextual Embedding Models
http://arxiv.org/abs/2004.02288
AUTHORS:        Subendhu Rongali ; Abhyuday Jagannatha ; Bhanu Pratap Singh Rawat ; Hong Yu
HIGHLIGHT:      We investigate methods to mitigate catastrophic forgetting during domain-specific pretraining of contextual embedding models such as BERT, DistilBERT, and RoBERTa.

69, TITLE:      Hierarchical Entity Typing via Multi-level Learning to Rank
http://arxiv.org/abs/2004.02286
AUTHORS:        Tongfei Chen ; Yunmo Chen ; Benjamin Van Durme
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      We propose a novel method for hierarchical entity classification that embraces ontological structure at both training and during prediction.

70, TITLE:      DualSDF: Semantic Shape Manipulation using a Two-Level Representation
http://arxiv.org/abs/2004.02869
AUTHORS:        Zekun Hao ; Hadar Averbuch-Elor ; Noah Snavely ; Serge Belongie
COMMENTS:       Published in CVPR 2020
HIGHLIGHT:      We propose DualSDF, a representation expressing shapes at two levels of granularity, one capturing fine details and the other representing an abstracted proxy shape using simple and semantically consistent shape primitives.

71, TITLE:      Game on Random Environement, Mean-field Langevin System and Neural Networks
http://arxiv.org/abs/2004.02457
AUTHORS:        Giovanni Conforti ; Anna Kazeykina ; Zhenjie Ren
HIGHLIGHT:      In this paper we study a type of games regularized by the relative entropy, where the players' strategies are coupled through a random environment variable.

72, TITLE:      Light Field Spatial Super-resolution via Deep Combinatorial Geometry Embedding and Structural Consistency Regularization
http://arxiv.org/abs/2004.02215
AUTHORS:        Jing Jin ; Junhui Hou ; Jie Chen ; Sam Kwong
COMMENTS:       This paper was accepted by CVPR 2020
HIGHLIGHT:      In this paper, we propose a novel learning-based LF spatial SR framework, in which each view of an LF image is first individually super-resolved by exploring the complementary information among views with combinatorial geometry embedding.

73, TITLE:      Lightweight Multi-View 3D Pose Estimation through Camera-Disentangled Representation
http://arxiv.org/abs/2004.02186
AUTHORS:        Edoardo Remelli ; Shangchen Han ; Sina Honari ; Pascal Fua ; Robert Wang
COMMENTS:       Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2020
HIGHLIGHT:      We present a lightweight solution to recover 3D pose from multi-view images captured with spatially calibrated cameras.

74, TITLE:      Iterative Context-Aware Graph Inference for Visual Dialog
http://arxiv.org/abs/2004.02194
AUTHORS:        Dan Guo ; Hui Wang ; Hanwang Zhang ; Zheng-Jun Zha ; Meng Wang
HIGHLIGHT:      To this end, we propose a novel Context-Aware Graph (CAG) neural network.

75, TITLE:      Deep Multimodal Feature Encoding for Video Ordering
http://arxiv.org/abs/2004.02205
AUTHORS:        Vivek Sharma ; Makarand Tapaswi ; Rainer Stiefelhagen
COMMENTS:       IEEE International Conference on Computer Vision (ICCV) Workshop on Large Scale Holistic Video Understanding. The datasets and code are available at https://github.com/vivoutlaw/tcbp
HIGHLIGHT:      We present a way to learn a compact multimodal feature representation that encodes all these modalities. To this end, we create a new multimodal dataset for temporal ordering that consists of approximately 30K scenes (2-6 clips per scene) based on the "Large Scale Movie Description Challenge".

76, TITLE:      Clustering based Contrastive Learning for Improving Face Representations
http://arxiv.org/abs/2004.02195
AUTHORS:        Vivek Sharma ; Makarand Tapaswi ; M. Saquib Sarfraz ; Rainer Stiefelhagen
COMMENTS:       To appear at IEEE International Conference on Automatic Face and Gesture Recognition (FG), 2020
HIGHLIGHT:      In this work, we present Clustering-based Contrastive Learning (CCL), a new clustering-based representation learning approach that uses labels obtained from clustering along with video constraints to learn discriminative face features.

77, TITLE:      Confident Coreset for Active Learning in Medical Image Analysis
http://arxiv.org/abs/2004.02200
AUTHORS:        Seong Tae Kim ; Farrukh Mushtaq ; Nassir Navab
COMMENTS:       10 pages, 4 figures
HIGHLIGHT:      In this paper, we propose a novel active learning method, confident coreset, which considers both uncertainty and distribution for effectively selecting informative samples.

78, TITLE:      Google Landmarks Dataset v2 -- A Large-Scale Benchmark for Instance-Level Recognition and Retrieval
http://arxiv.org/abs/2004.01804
AUTHORS:        Tobias Weyand ; Andre Araujo ; Bingyi Cao ; Jack Sim
COMMENTS:       CVPR20 camera-ready (oral)
HIGHLIGHT:      We introduce the Google Landmarks Dataset v2 (GLDv2), a new benchmark for large-scale, fine-grained instance recognition and image retrieval in the domain of human-made and natural landmarks.

79, TITLE:      Self-Supervised Viewpoint Learning From Image Collections
http://arxiv.org/abs/2004.01793
AUTHORS:        Siva Karthik Mustikovela ; Varun Jampani ; Shalini De Mello ; Sifei Liu ; Umar Iqbal ; Carsten Rother ; Jan Kautz
COMMENTS:       Accepted at CVPR 20
HIGHLIGHT:      We propose a novel learning framework which incorporates an analysis-by-synthesis paradigm to reconstruct images in a viewpoint aware manner with a generative network, along with symmetry and adversarial constraints to successfully supervise our viewpoint estimation network.

80, TITLE:      SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation
http://arxiv.org/abs/2004.01803
AUTHORS:        Chenfeng Xu ; Bichen Wu ; Zining Wang ; Wei Zhan ; Peter Vajda ; Kurt Keutzer ; Masayoshi Tomizuka
COMMENTS:       Code and data are available at: https://github.com/chenfengxu714/SqueezeSegV3.git
HIGHLIGHT:      To fix this, we propose Spatially-Adaptive Convolution (SAC) to adopt different filters for different locations according to the input image.

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

82, TITLE:      TimeGate: Conditional Gating of Segments in Long-range Activities
http://arxiv.org/abs/2004.01808
AUTHORS:        Noureldien Hussein ; Mihir Jain ; Babak Ehteshami Bejnordi
HIGHLIGHT:      We propose TimeGate, along with a novel conditional gating module, for sampling the most representative segments from the long-range activity.

83, TITLE:      Privacy-Preserving Eye Videos using Rubber Sheet Model
http://arxiv.org/abs/2004.01792
AUTHORS:        Aayush K Chaudhary ; Jeff B. Pelz
COMMENTS:       Will be published in ETRA 20 Short Papers, June 2-5, 2020, Stuttgart, Germany Copyright 2020 Association for Computing Machinery
HIGHLIGHT:      We present a new approach to handle privacy issues in eye videos by replacing the current identifiable iris texture with a different iris template in the video capture pipeline based on the Rubber Sheet Model.

84, TITLE:      The Collection Virtual Machine: An Abstraction for Multi-Frontend Multi-Backend Data Analysis
http://arxiv.org/abs/2004.01908
AUTHORS:        Sabir Akhadov ; Renato Marroquín ; Ingo Müller
COMMENTS:       This paper was originally submitted to CIDR'19
HIGHLIGHT:      In this paper, we propose the "Collection Virtual Machine" (or CVM) -- an abstraction that can capture the essence of a large span of types of analyses and backend implementations.

85, TITLE:      A Generic Graph-based Neural Architecture Encoding Scheme for Predictor-based NAS
http://arxiv.org/abs/2004.01899
AUTHORS:        Xuefei Ning ; Yin Zheng ; Tianchen Zhao ; Yu Wang ; Huazhong Yang
COMMENTS:       15 pages main text; 9 pages appendix
HIGHLIGHT:      This work proposes a novel Graph-based neural ArchiTecture Encoding Scheme, a.k.a. GATES, to improve the predictor-based neural architecture search.

86, TITLE:      Complex-Valued Convolutional Neural Networks for MRI Reconstruction
http://arxiv.org/abs/2004.01738
AUTHORS:        Elizabeth K. Cole ; Joseph Y. Cheng ; John M. Pauly ; Shreyas S. Vasanawala
COMMENTS:       10 pages; 11 figures; Submitted to IEEE TMI
HIGHLIGHT:      In this work, we investigate end-to-end complex-valued convolutional neural networks - specifically, for image reconstruction in lieu of two-channel real-valued networks.

87, TITLE:      A limitation on the KPT interpolation
http://arxiv.org/abs/2004.02448
AUTHORS:        Jan Krajicek
COMMENTS:       6 pages, preprint March 2020
HIGHLIGHT:      We prove a limitation on a variant of the KPT theorem proposed for propositional proof systems by Pich and Santhanam 2020, for all proof systems that prove the disjointness of two NP sets that are hard to distinguish.

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

89, TITLE:      Sparse Text Generation
http://arxiv.org/abs/2004.02644
AUTHORS:        Pedro Henrique Martins ; Zita Marinho ; André F. T. Martins
HIGHLIGHT:      In this paper, we use the recently introduced entmax transformation to train and sample from a natively sparse language model, avoiding this mismatch.

90, TITLE:      At Which Level Should We Extract? An Empirical Study on Extractive Document Summarization
http://arxiv.org/abs/2004.02664
AUTHORS:        Qingyu Zhou ; Furu Wei ; Ming Zhou
HIGHLIGHT:      In this work, we show that unnecessity and redundancy issues exist when extracting full sentences, and extracting sub-sentential units is a promising alternative.

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

92, TITLE:      Learning to Summarize Passages: Mining Passage-Summary Pairs from Wikipedia Revision Histories
http://arxiv.org/abs/2004.02592
AUTHORS:        Qingyu Zhou ; Furu Wei ; Ming Zhou
HIGHLIGHT:      In this paper, we propose a method for automatically constructing a passage-to-summary dataset by mining the Wikipedia page revision histories.

93, TITLE:      Graph Sequential Network for Reasoning over Sequences
http://arxiv.org/abs/2004.02001
AUTHORS:        Ming Tu ; Jing Huang ; Xiaodong He ; Bowen Zhou
COMMENTS:       Part of this paper was presented at NeurIPS 2019 Workshop on Graph Representation Learning
HIGHLIGHT:      In this paper, we consider a novel case where reasoning is needed over graphs built from sequences, i.e. graph nodes with sequence data.

94, TITLE:      Hooks in the Headline: Learning to Generate Headlines with Controlled Styles
http://arxiv.org/abs/2004.01980
AUTHORS:        Di Jin ; Zhijing Jin ; Joey Tianyi Zhou ; Lisa Orii ; Peter Szolovits
COMMENTS:       To appear at ACL 2020
HIGHLIGHT:      We propose a new task, Stylistic Headline Generation (SHG), to enrich the headlines with three style options (humor, romance and clickbait), in order to attract more readers.

95, TITLE:      Generating Hierarchical Explanations on Text Classification via Feature Interaction Detection
http://arxiv.org/abs/2004.02015
AUTHORS:        Hanjie Chen ; Guangtao Zheng ; Yangfeng Ji
COMMENTS:       Accepted to ACL 2020
HIGHLIGHT:      In this work, we build hierarchical explanations by detecting feature interactions.

96, TITLE:      Learning a Simple and Effective Model for Multi-turn Response Generation with Auxiliary Tasks
http://arxiv.org/abs/2004.01972
AUTHORS:        Yufan Zhao ; Can Xu ; Wei Wu
HIGHLIGHT:      In this work, we pursue a model that has a simple structure yet can effectively leverage conversation contexts for response generation.

97, TITLE:      Open Domain Dialogue Generation with Latent Images
http://arxiv.org/abs/2004.01981
AUTHORS:        Ze Yang ; Wei Wu ; Huang Hu ; Can Xu ; Zhoujun Li
HIGHLIGHT:      Thus, we propose learning a response generation model with both image-grounded dialogues and textual dialogues by assuming that there is a latent variable in a textual dialogue that represents the image, and trying to recover the latent image through text-to-image generation techniques.

98, TITLE:      End-to-End Abstractive Summarization for Meetings
http://arxiv.org/abs/2004.02016
AUTHORS:        Chenguang Zhu ; Ruochen Xu ; Michael Zeng ; Xuedong Huang
COMMENTS:       12 pages, 2 figures
HIGHLIGHT:      In this paper, we propose a novel end-to-end abstractive summary network that adapts to the meeting scenario.

99, TITLE:      Incorporating Bilingual Dictionaries for Low Resource Semi-Supervised Neural Machine Translation
http://arxiv.org/abs/2004.02071
AUTHORS:        Sreyashi Nag ; Mihir Kale ; Varun Lakshminarasimhan ; Swapnil Singhavi
HIGHLIGHT:      We propose a simple data augmentation technique to address both this shortcoming.

100, TITLE:     Talk to Papers: Bringing Neural Question Answering to Academic Search
http://arxiv.org/abs/2004.02002
AUTHORS:        Tianchang Zhao ; Kyusong Lee
COMMENTS:       accepted at ACL 2020, demo paper
HIGHLIGHT:      We introduce Talk to Papers, which exploits the recent open-domain question answering (QA) techniques to improve the current experience of academic search.

101, TITLE:     Geometrically Principled Connections in Graph Neural Networks
http://arxiv.org/abs/2004.02658
AUTHORS:        Shunwang Gong ; Mehdi Bahri ; Michael M. Bronstein ; Stefanos Zafeiriou
COMMENTS:       Presented at Computer Vision and Pattern Recognition (CVPR), 2020
HIGHLIGHT:      In this paper, we argue geometry should remain the primary driving force behind innovation in the emerging field of geometric deep learning.

102, TITLE:     Appearance Shock Grammar for Fast Medial Axis Extraction from Real Images
http://arxiv.org/abs/2004.02677
AUTHORS:        Charles-Olivier Dufresne Camaro ; Morteza Rezanejad ; Stavros Tsogkas ; Kaleem Siddiqi ; Sven Dickinson
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      We combine ideas from shock graph theory with more recent appearance-based methods for medial axis extraction from complex natural scenes, improving upon the present best unsupervised method, in terms of efficiency and performance.

103, TITLE:     Finding Your (3D) Center: 3D Object Detection Using a Learned Loss
http://arxiv.org/abs/2004.02693
AUTHORS:        David Griffiths ; Jan Boehm ; Tobias Ritschel
COMMENTS:       19 pages, 9 figures
HIGHLIGHT:      Addressing this disparity, we introduce a new optimization procedure, which allows training for 3D detection with raw 3D scans while using as little as 5% of the object labels and still achieve comparable performance.

104, TITLE:     Light3DPose: Real-time Multi-Person 3D PoseEstimation from Multiple Views
http://arxiv.org/abs/2004.02688
AUTHORS:        Alessio Elmi ; Davide Mazzini ; Pietro Tortella
HIGHLIGHT:      We present an approach to perform 3D pose estimation of multiple people from a few calibrated camera views.

105, TITLE:     SHOP-VRB: A Visual Reasoning Benchmark for Object Perception
http://arxiv.org/abs/2004.02673
AUTHORS:        Michal Nazarczuk ; Krystian Mikolajczyk
COMMENTS:       International Conference on Robotics and Automation (ICRA) 2020
HIGHLIGHT:      In this paper we present an approach and a benchmark for visual reasoning in robotics applications, in particular small object grasping and manipulation.

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

107, TITLE:     Attribute Mix: Semantic Data Augmentation for Fine Grained Recognition
http://arxiv.org/abs/2004.02684
AUTHORS:        Hao Li ; Xiaopeng Zhang ; Hongkai Xiong ; Qi Tian
HIGHLIGHT:      In this paper, we propose Attribute Mix, a data augmentation strategy at attribute level to expand the fine-grained samples.

108, TITLE:     Fair Latency-Aware Metric for real-time video segmentation networks
http://arxiv.org/abs/2004.02574
AUTHORS:        Evann Courdier ; Francois Fleuret
HIGHLIGHT:      In this paper, we argue that this metric is not relevant enough for real-time video as it does not take into account the processing time (latency) of the network.

109, TITLE:     Anisotropic Convolutional Networks for 3D Semantic Scene Completion
http://arxiv.org/abs/2004.02122
AUTHORS:        Jie Li ; Kai Han ; Peng Wang ; Yu Liu ; Xia Yuan
HIGHLIGHT:      To handle such variations, we propose a novel module called anisotropic convolution, which properties with flexibility and power impossible for the competing methods such as standard 3D convolution and some of its variations.

110, TITLE:     Neuron Linear Transformation: Modeling the Domain Shift for Crowd Counting
http://arxiv.org/abs/2004.02133
AUTHORS:        Qi Wang ; Tao Han ; Junyu Gao ; Yuan Yuan
COMMENTS:       12 paegs, 8 figures
HIGHLIGHT:      To describe the domain gap directly at the parameter-level, we propose a Neuron Linear Transformation (NLT) method, where NLT is exploited to learn the shift at neuron-level and then transfer the source model to the target model.

111, TITLE:     Learning and Recognizing Archeological Features from LiDAR Data
http://arxiv.org/abs/2004.02099
AUTHORS:        Conrad M Albrecht ; Chris Fisher ; Marcus Freitag ; Hendrik F Hamann ; Sharathchandra Pankanti ; Florencia Pezzutti ; Francesca Rossi
HIGHLIGHT:      We present a remote sensing pipeline that processes LiDAR (Light Detection And Ranging) data through machine & deep learning for the application of archeological feature detection on big geo-spatial data platforms such as e.g. IBM PAIRS Geoscope.

112, TITLE:     Deep Homography Estimation for Dynamic Scenes
http://arxiv.org/abs/2004.02132
AUTHORS:        Hoang Le ; Feng Liu ; Shu Zhang ; Aseem Agarwala
COMMENTS:       CVPR 2020, https://github.com/lcmhoang/hmg-dynamics
HIGHLIGHT:      This paper investigates and discusses how to design and train a deep neural network that handles dynamic scenes. We first collect a large video dataset with dynamic content.

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

114, TITLE:     Deeply Aligned Adaptation for Cross-domain Object Detection
http://arxiv.org/abs/2004.02093
AUTHORS:        Minghao Fu ; Zhenshan Xie ; Wen Li ; Lixin Duan
HIGHLIGHT:      In this work, we propose an end-to-end solution based on Faster R-CNN, where ground-truth annotations are available for source images (e.g., cartoon) but not for target ones (e.g., watercolor) during training.

115, TITLE:     TAPAS: Weakly Supervised Table Parsing via Pre-training
http://arxiv.org/abs/2004.02349
AUTHORS:        Jonathan Herzig ; Paweł Krzysztof Nowak ; Thomas Müller ; Francesco Piccinno ; Julian Martin Eisenschlos
COMMENTS:       Accepted to ACL 2020
HIGHLIGHT:      In this paper, we present TAPAS, an approach to question answering over tables without generating logical forms.

116, TITLE:     Measuring Social Biases of Crowd Workers using Counterfactual Queries
http://arxiv.org/abs/2004.02028
AUTHORS:        Bhavya Ghai ; Q. Vera Liao ; Yunfeng Zhang ; Klaus Mueller
COMMENTS:       Accepted at the Workshop on Fair and Responsible AI at ACM CHI 2020
HIGHLIGHT:      In this work, we propose a new method based on counterfactual fairness to quantify the degree of inherent social bias in each crowd worker.

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

118, TITLE:     XtracTree for Regulator Validation of Bagging Methods Used in Retail Banking
http://arxiv.org/abs/2004.02326
AUTHORS:        Jeremy Charlier ; Vladimir Makarenkov
HIGHLIGHT:      In this context, we propose XtracTree, an algorithm that is capable of effectively converting an ML bagging classifier, such as a decision tree or a random forest, into simple "if-then" rules satisfying the requirements of model validation.

119, TITLE:     Personalization in Human-AI Teams: Improving the Compatibility-Accuracy Tradeoff
http://arxiv.org/abs/2004.02289
AUTHORS:        Jonathan Martinez ; Kobi Gal ; Ece Kamar ; Levi H. S. Lelis
COMMENTS:       8 pages, 7 figures
HIGHLIGHT:      In this paper, we show that in some cases it is possible to improve this compatibility-accuracy trade-off relative to a specific user by employing new error functions for the AI updates that personalize the weight updates to be compatible with the user's history of interaction with the system and present experimental results indicating that this approach provides major improvements to certain users.

120, TITLE:     Dynamic Decision Boundary for One-class Classifiers applied to non-uniformly Sampled Data
http://arxiv.org/abs/2004.02273
AUTHORS:        Riccardo La Grassa ; Ignazio Gallo ; Nicola Landro
COMMENTS:       7 pages
HIGHLIGHT:      In this paper, we propose a one-class classifier based on the minimum spanning tree with a dynamic decision boundary (OCdmst) to make good prediction also in the case we have non-uniformly sampled data.

121, TITLE:     An Eigenspace Divide-and-Conquer Approach for Large-Scale Optimization
http://arxiv.org/abs/2004.02115
AUTHORS:        Zhigang Ren ; Yongsheng Liang ; Muyi Wang ; Yang Yang ; An Chen
HIGHLIGHT:      This study attempts to address the above issue from a different perspective and proposes an eigenspace divide-and-conquer (EDC) approach.

122, TITLE:     Responsive Parallelism with Futures and State
http://arxiv.org/abs/2004.02870
AUTHORS:        Stefan K. Muller ; Kyle Singer ; Noah Goldstein ; Umut A. Acar ; Kunal Agrawal ; I-Ting Angelina Lee
HIGHLIGHT:      In this paper, we lift the restriction concerning the use of state.

123, TITLE:     Bringing GNU Emacs to Native Code
http://arxiv.org/abs/2004.02504
AUTHORS:        Andrea Corallo ; Luca Nassi ; Nicola Manca
COMMENTS:       8 pages, 1 figure
HIGHLIGHT:      In this work we discuss the implementation of an optimizing compiler approach for Elisp targeting native code.

124, TITLE:     Rational neural networks
http://arxiv.org/abs/2004.01902
AUTHORS:        Nicolas Boullé ; Yuji Nakatsukasa ; Alex Townsend
COMMENTS:       20 pages, 6 figures
HIGHLIGHT:      We consider neural networks with rational activation functions.

125, TITLE:     Vocoder-Based Speech Synthesis from Silent Videos
http://arxiv.org/abs/2004.02541
AUTHORS:        Daniel Michelsanti ; Olga Slizovskaia ; Gloria Haro ; Emilia Gómez ; Zheng-Hua Tan ; Jesper Jensen
HIGHLIGHT:      In this paper, we present a way to synthesise speech from the silent video of a talker using deep learning.

126, TITLE:     Evaluating NLP Models via Contrast Sets
http://arxiv.org/abs/2004.02709
AUTHORS:        Matt Gardner ; Yoav Artzi ; Victoria Basmova ; Jonathan Berant ; Ben Bogin ; Sihao Chen ; Pradeep Dasigi ; Dheeru Dua ; Yanai Elazar ; Ananth Gottumukkala ; Nitish Gupta ; Hanna Hajishirzi ; Gabriel Ilharco ; Daniel Khashabi ; Kevin Lin ; Jiangming Liu ; Nelson F. Liu ; Phoebe Mulcaire ; Qiang Ning ; Sameer Singh ; Noah A. Smith ; Sanjay Subramanian ; Reut Tsarfaty ; Eric Wallace ; Ally Zhang ; Ben Zhou
HIGHLIGHT:      We propose a new annotation paradigm for NLP that helps to close systematic gaps in the test data.

127, TITLE:     Leveraging the Inherent Hierarchy of Vacancy Titles for Automated Job Ontology Expansion
http://arxiv.org/abs/2004.02814
AUTHORS:        Jeroen Van Hautte ; Vincent Schelstraete ; Mikaël Wornoo
COMMENTS:       Accepted to the Proceedings of the 6th International Workshop on Computational Terminology (COMPUTERM 2020)
HIGHLIGHT:      We introduce a novel, purely data-driven approach towards the detection of new job titles.

128, TITLE:     COVID-19: Strategies for Allocation of Test Kits
http://arxiv.org/abs/2004.01740
AUTHORS:        Arpita Biswas ; Shruthi Bannur ; Prateek Jain ; Srujana Merugu
HIGHLIGHT:      In this report, we consider the problem of allocating test-kits and discuss some solution approaches.

129, TITLE:     Generative Forensics: Procedural Generation and Information Games
http://arxiv.org/abs/2004.01768
AUTHORS:        Michael Cook
HIGHLIGHT:      In this paper we introduce the notion of generative forensics games, a subgenre of information games that challenge the player to understand the output of a generative system.

130, TITLE:     Meta-Learning for Few-Shot NMT Adaptation
http://arxiv.org/abs/2004.02745
AUTHORS:        Amr Sharaf ; Hany Hassan ; Hal Daumé III
HIGHLIGHT:      We present META-MT, a meta-learning approach to adapt Neural Machine Translation (NMT) systems in a few-shot setting.

131, TITLE:     Quantum Inspired Word Representation and Computation
http://arxiv.org/abs/2004.02705
AUTHORS:        Shen Li ; Renfen Hu ; Jinshan Wu
HIGHLIGHT:      Inspired by quantum probability, we represent words as density matrices, which are inherently capable of representing mixed states.

132, TITLE:     GIANT: Scalable Creation of a Web-scale Ontology
http://arxiv.org/abs/2004.02118
AUTHORS:        Bang Liu ; Weidong Guo ; Di Niu ; Jinwen Luo ; Chaoyue Wang ; Zhen Wen ; Yu Xu
COMMENTS:       Accepted as full paper by SIGMOD 2020
HIGHLIGHT:      In this paper, we present GIANT, a mechanism to construct a user-centered, web-scale, structured ontology, containing a large number of natural language phrases conforming to user attentions at various granularities, mined from a vast volume of web documents and search click graphs.

133, TITLE:     Reference Language based Unsupervised Neural Machine Translation
http://arxiv.org/abs/2004.02127
AUTHORS:        Zuchao Li ; Hai Zhao ; Rui Wang ; Masao Utiyama ; Eiichiro Sumita
HIGHLIGHT:      Experimental results show that our methods improve the quality of UNMT over that of a strong baseline in terms of only one auxiliary language, demonstrating the usefulness of the proposed reference language based UNMT with a good start.

134, TITLE:     Unsupervised Domain Clusters in Pretrained Language Models
http://arxiv.org/abs/2004.02105
AUTHORS:        Roee Aharoni ; Yoav Goldberg
COMMENTS:       Accepted as a long paper in ACL 2020
HIGHLIGHT:      We harness this property and propose domain data selection methods based on such models, which require only a small set of in-domain monolingual data.

135, TITLE:     Machine Translation Pre-training for Data-to-Text Generation -- A Case Study in Czech
http://arxiv.org/abs/2004.02077
AUTHORS:        Mihir Kale ; Scott Roy
HIGHLIGHT:      In this paper, we study the effectiveness of machine translation based pre-training for data-to-text generation in non-English languages.

136, TITLE:     A Resource for Studying Chatino Verbal Morphology
http://arxiv.org/abs/2004.02083
AUTHORS:        Hilaria Cruz ; Gregory Stump ; Antonios Anastasopoulos
COMMENTS:       accepted at LREC 2020
HIGHLIGHT:      We present the first resource focusing on the verbal inflectional morphology of San Juan Quiahije Chatino, a tonal mesoamerican language spoken in Mexico.

137, TITLE:     COVID-CAPS: A Capsule Network-based Framework for Identification of COVID-19 cases from X-ray Images
http://arxiv.org/abs/2004.02696
AUTHORS:        Parnian Afshar ; Shahin Heidarian ; Farnoosh Naderkhani ; Anastasia Oikonomou ; Konstantinos N. Plataniotis ; Arash Mohammadi
HIGHLIGHT:      This paper presents an alternative modeling framework based on Capsule Networks, referred to as the COVID-CAPS, being capable of handling small datasets, which is of significant importance due to sudden and rapid emergence of COVID-19.

138, TITLE:     Network Adjustment: Channel Search Guided by FLOPs Utilization Ratio
http://arxiv.org/abs/2004.02767
AUTHORS:        Zhengsu Chen ; Jianwei Niu ; Lingxi Xie ; Xuefeng Liu ; Longhui Wei ; Qi Tian
HIGHLIGHT:      This paper presents a new framework named network adjustment, which considers network accuracy as a function of FLOPs, so that under each network configuration, one can estimate the FLOPs utilization ratio (FUR) for each layer and use it to determine whether to increase or decrease the number of channels on the layer.

139, TITLE:     SSN: Shape Signature Networks for Multi-class Object Detection from Point Clouds
http://arxiv.org/abs/2004.02774
AUTHORS:        Xinge Zhu ; Yuexin Ma ; Tai Wang ; Yan Xu ; Jianping Shi ; Dahua Lin
COMMENTS:       Code is available at https://github.com/xinge008/SSN
HIGHLIGHT:      In this paper, we propose a novel 3D shape signature to explore the shape information from point clouds.

140, TITLE:     A Morphable Face Albedo Model
http://arxiv.org/abs/2004.02711
AUTHORS:        William A. P. Smith ; Alassane Seck ; Hannah Dee ; Bernard Tiddeman ; Joshua Tenenbaum ; Bernhard Egger
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we bring together two divergent strands of research: photometric face capture and statistical 3D face appearance modelling.

141, TITLE:     Sub-Instruction Aware Vision-and-Language Navigation
http://arxiv.org/abs/2004.02707
AUTHORS:        Yicong Hong ; Cristian Rodriguez-Opazo ; Qi Wu ; Stephen Gould
HIGHLIGHT:      In this work, we focus on the granularity of the visual and language sequences as well as the trackability of agents through the completion of instruction.

142, TITLE:     Reconfigurable Voxels: A New Representation for LiDAR-Based Point Clouds
http://arxiv.org/abs/2004.02724
AUTHORS:        Tai Wang ; Xinge Zhu ; Dahua Lin
HIGHLIGHT:      To tackle this difficulty, we propose Reconfigurable Voxels, a new approach to constructing representations from 3D point clouds.

143, TITLE:     Guiding Monocular Depth Estimation Using Depth-Attention Volume
http://arxiv.org/abs/2004.02760
AUTHORS:        Lam Huynh ; Phong Nguyen-Ha ; Jiri Matas ; Esa Rahtu ; Janne Heikkila
HIGHLIGHT:      In this paper, we propose guiding depth estimation to favor planar structures that are ubiquitous especially in indoor environments.

144, TITLE:     BiSeNet V2: Bilateral Network with Guided Aggregation for Real-time Semantic Segmentation
http://arxiv.org/abs/2004.02147
AUTHORS:        Changqian Yu ; Changxin Gao ; Jingbo Wang ; Gang Yu ; Chunhua Shen ; Nong Sang
COMMENTS:       16 pages, 10 figures, 9 tables. Code is available at https://git.io/BiSeNet
HIGHLIGHT:      We propose to treat these spatial details and categorical semantics separately to achieve high accuracy and high efficiency for realtime semantic segmentation.

145, TITLE:     Comparative Analysis of Multiple Deep CNN Models for Waste Classification
http://arxiv.org/abs/2004.02168
AUTHORS:        Dipesh Gyawali ; Alok Regmi ; Aatish Shakya ; Ashish Gautam ; Surendra Shrestha
COMMENTS:       6 pages, 13 figures
HIGHLIGHT:      Without the human exercise in segregating those waste products, the study would save the precious time and would introduce the automation in the area of waste management.

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

147, TITLE:     Flow2Stereo: Effective Self-Supervised Learning of Optical Flow and Stereo Matching
http://arxiv.org/abs/2004.02138
AUTHORS:        Pengpeng Liu ; Irwin King ; Michael Lyu ; Jia Xu
COMMENTS:       Published at the Conference on Computer Vision and Pattern Recognition (CVPR 2020)
HIGHLIGHT:      In this paper, we propose a unified method to jointly learn optical flow and stereo matching.

148, TITLE:     Adversarial-Prediction Guided Multi-task Adaptation for Semantic Segmentation of Electron Microscopy Images
http://arxiv.org/abs/2004.02134
AUTHORS:        Jiajin Yi ; Zhimin Yuan ; Jialin Peng
HIGHLIGHT:      In this study, we introduce an adversarial-prediction guided multi-task network to learn the adaptation of a well-trained model for use on a novel unlabeled target domain.

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

150, TITLE:     #P-completeness of counting update digraphs, cacti, and a series-parallel decomposition method
http://arxiv.org/abs/2004.02129
AUTHORS:        Camille Noûs ; Kévin Perrot ; Sylvain Sené ; Lucas Venturini
HIGHLIGHT:      We prove that counting the number of equivalence classes, that is a tight upper bound on the synchronism sensitivity of a given network, is #P-complete.

151, TITLE:     Using Generative Adversarial Nets on Atari Games for Feature Extraction in Deep Reinforcement Learning
http://arxiv.org/abs/2004.02762
AUTHORS:        Ayberk Aydın ; Elif Surer
COMMENTS:       in Turkish
HIGHLIGHT:      In this study, Proximal Policy Optimization (PPO) algorithm is augmented with Generative Adversarial Networks (GANs) to increase the sample efficiency by enforcing the network to learn efficient representations without depending on sparse and delayed rewards as supervision.

152, TITLE:     Adaptive Partial Scanning Transmission Electron Microscopy with Reinforcement Learning
http://arxiv.org/abs/2004.02786
AUTHORS:        Jeffrey M. Ede
COMMENTS:       12 pages, 4 figures. Appendix: 8 pages
HIGHLIGHT:      We have extended recurrent deterministic policy gradients to train deep LSTMs and differentiable neural computers to adaptively sample scan path segments.

153, TITLE:     A Norm Emergence Framework for Normative MAS -- Position Paper
http://arxiv.org/abs/2004.02575
AUTHORS:        Andreasa Morris-Martin ; Marina De Vos ; Julian Padget
COMMENTS:       16 pages, 2 figures, pre-print for International Workshop on Coordination, Organizations, Institutions, Norms and Ethics for Governance of Multi-Agent Systems (COINE), co-located with AAMAS 2020
HIGHLIGHT:      In this paper, we make the case that, similarly, a norm has emerged in a normative MAS when a percentage of agents adopt the norm.

154, TITLE:     Networked Multi-Agent Reinforcement Learning with Emergent Communication
http://arxiv.org/abs/2004.02780
AUTHORS:        Shubham Gupta ; Rishi Hazra ; Ambedkar Dukkipati
COMMENTS:       An abridged version of this paper has been accepted as a short paper at AAMAS 2020
HIGHLIGHT:      In this paper, we formulate and study a MARL problem where cooperative agents are connected to each other via a fixed underlying network.

155, TITLE:     Emotional Video to Audio Transformation Using Deep Recurrent Neural Networks and a Neuro-Fuzzy System
http://arxiv.org/abs/2004.02113
AUTHORS:        Gwenaelle Cunha Sergio ; Minho Lee
COMMENTS:       Published (https://www.hindawi.com/journals/mpe/2020/8478527/)
HIGHLIGHT:      In this study, we propose a novel hybrid deep neural network that uses an Adaptive Neuro-Fuzzy Inference System to predict a video's emotion from its visual features and a deep Long Short-Term Memory Recurrent Neural Network to generate its corresponding audio signals with similar emotional inkling.

156, TITLE:     Adaptive Explainable Neural Networks (AxNNs)
http://arxiv.org/abs/2004.02353
AUTHORS:        Jie Chen ; Joel Vaughan ; Vijayan N. Nair ; Agus Sudjianto
HIGHLIGHT:      We develop a new framework called Adaptive Explainable Neural Networks (AxNN) for achieving the dual goals of good predictive performance and model interpretability.

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

158, TITLE:     Coronavirus Detection and Analysis on Chest CT with Deep Learning
http://arxiv.org/abs/2004.02640
AUTHORS:        Ophir Gozes ; Maayan Frid-Adar ; Nimrod Sagie ; Huangqi Zhang ; Wenbin Ji ; Hayit Greenspan
HIGHLIGHT:      We present our results on a dataset comprised of 110 confirmed COVID-19 patients from Zhejiang province, China.

159, TITLE:     Automatic Right Ventricle Segmentation using Multi-Label Fusion in Cardiac MRI
http://arxiv.org/abs/2004.02317
AUTHORS:        Maria A. Zuluaga ; M. Jorge Cardoso ; Sébastien Ourselin
HIGHLIGHT:      This paper presents a fully automatic method for the segmentation of the RV in cardiac magnetic resonance images (MRI).

160, TITLE:     Lossless Image Compression through Super-Resolution
http://arxiv.org/abs/2004.02872
AUTHORS:        Sheng Cao ; Chao-Yuan Wu ; Philipp Krähenbühl
COMMENTS:       Tech report
HIGHLIGHT:      We introduce a simple and efficient lossless image compression algorithm.

161, TITLE:     Domain-based Latent Personal Analysis and its use for impersonation detection in social media
http://arxiv.org/abs/2004.02346
AUTHORS:        Osnat Mokryn ; Hagit Ben-Shoshan
COMMENTS:       Submitted to Journal
HIGHLIGHT:      We devise a method, termed Latent Personal Analysis (LPA), for finding such domain-based personal signatures.

162, TITLE:     BERT in Negotiations: Early Prediction of Buyer-Seller Negotiation Outcomes
http://arxiv.org/abs/2004.02363
AUTHORS:        Kushal Chawla ; Gale Lucas ; Jonathan Gratch ; Jonathan May
HIGHLIGHT:      Towards this end, we aim to understand the role of natural language in negotiations from a data-driven perspective by attempting to predict a negotiation's outcome, well before the negotiation is complete.

163, TITLE:     Learning to Recover Reasoning Chains for Multi-Hop Question Answering via Cooperative Games
http://arxiv.org/abs/2004.02393
AUTHORS:        Yufei Feng ; Mo Yu ; Wenhan Xiong ; Xiaoxiao Guo ; Junjie Huang ; Shiyu Chang ; Murray Campbell ; Michael Greenspan ; Xiaodan Zhu
HIGHLIGHT:      We propose a cooperative game approach to deal with this problem, in which how the evidence passages are selected and how the selected passages are connected are handled by two models that cooperate to select the most confident chains from a large set of candidates (from distant supervision). We propose the new problem of learning to recover reasoning chains from weakly supervised signals, i.e., the question-answer pairs. For evaluation, we created benchmarks based on two multi-hop QA datasets, HotpotQA and MedHop; and hand-labeled reasoning chains for the latter.

164, TITLE:     Neural Machine Translation with Imbalanced Classes
http://arxiv.org/abs/2004.02334
AUTHORS:        Thamme Gowda ; Jonathan May
HIGHLIGHT:      We cast neural machine translation (NMT) as a classification task in an autoregressive setting and analyze the limitations of both classification and autoregression components.

165, TITLE:     Feature Super-Resolution Based Facial Expression Recognition for Multi-scale Low-Resolution Faces
http://arxiv.org/abs/2004.02234
AUTHORS:        Wei Jing ; Feng Tian ; Jizhong Zhang ; Kuo-Ming Chao ; Zhenxin Hong ; Xu Liu
COMMENTS:       13 pages, 5 figures
HIGHLIGHT:      In this work, inspired by feature super-resolution methods for object detection, we proposed a novel generative adversary network-based feature level super-resolution method for robust facial expression recognition(FSR-FER).

166, TITLE:     EfficientPS: Efficient Panoptic Segmentation
http://arxiv.org/abs/2004.02307
AUTHORS:        Rohit Mohan ; Abhinav Valada
COMMENTS:       A live demo is available at http://panoptic.cs.uni-freiburg.de and the code as well as the models are available at https://github.com/DeepSceneSeg/EfficientPS
HIGHLIGHT:      In this paper, we introduce the Efficient Panoptic Segmentation (EfficientPS) architecture that consists of a shared backbone which efficiently encodes and fuses semantically rich multi-scale features. Additionally, we introduce the KITTI panoptic segmentation dataset that contains panoptic annotations for the popularly challenging KITTI benchmark.

167, TITLE:     Steering Self-Supervised Feature Learning Beyond Local Pixel Statistics
http://arxiv.org/abs/2004.02331
AUTHORS:        Simon Jenni ; Hailin Jin ; Paolo Favaro
COMMENTS:       CVPR 2020 (oral)
HIGHLIGHT:      We introduce a novel principle for self-supervised feature learning based on the discrimination of specific transformations of an image.

168, TITLE:     Hyper-spectral NIR and MIR data and optimal wavebands for detecting of apple trees diseases
http://arxiv.org/abs/2004.02325
AUTHORS:        Dmitrii Shadrin ; Mariia Pukalchik ; Anastasia Uryasheva ; Nikita Rodichenko ; Dzmitry Tsetserukou
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      This research proposes a modern approach for analysing spectral data in Near-Infrared and Mid-Infrared ranges of the apple trees diseases on different stages.

169, TITLE:     Structural-analogy from a Single Image Pair
http://arxiv.org/abs/2004.02222
AUTHORS:        Sagie Benaim ; Ron Mokady ; Amit Bermano ; Daniel Cohen-Or ; Lior Wolf
HIGHLIGHT:      In this paper, we explore the capabilities of neural networks to understand image structure given only a single pair of images, A and B.

170, TITLE:     Group Based Deep Shared Feature Learning for Fine-grained Image Classification
http://arxiv.org/abs/2004.01817
AUTHORS:        Xuelu Li ; Vishal Monga
HIGHLIGHT:      Given that images from distinct classes in fine-grained classification share significant features of interest, we present a new deep network architecture that explicitly models shared features and removes their effect to achieve enhanced classification results.

171, TITLE:     Deblurring by Realistic Blurring
http://arxiv.org/abs/2004.01860
AUTHORS:        Kaihao Zhang ; Wenhan Luo ; Yiran Zhong ; Lin Ma ; Bjorn Stenger ; Wei Liu ; Hongdong Li
HIGHLIGHT:      To address this problem, we propose a new method which combines two GAN models, i.e., a learning-to-Blur GAN (BGAN) and learning-to-DeBlur GAN (DBGAN), in order to learn a better model for image deblurring by primarily learning how to blur images. As an additional contribution, this paper also introduces a Real-World Blurred Image (RWBI) dataset including diverse blurry images.

172, TITLE:     Pixel Consensus Voting for Panoptic Segmentation
http://arxiv.org/abs/2004.01849
AUTHORS:        Haochen Wang ; Ruotian Luo ; Michael Maire ; Greg Shakhnarovich
COMMENTS:       CVPR 2020
HIGHLIGHT:      We implement vote aggregation and backprojection using native operators of a convolutional neural network.

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

174, TITLE:     Multi-Variate Temporal GAN for Large Scale Video Generation
http://arxiv.org/abs/2004.01823
AUTHORS:        Andres Muñoz ; Mohammadreza Zolfaghari ; Max Argus ; Thomas Brox
COMMENTS:       13 pages, 13 figures
HIGHLIGHT:      In this paper, we present a network architecture for video generation that models spatio-temporal consistency without resorting to costly 3D architectures.

175, TITLE:     Feature Quantization Improves GAN Training
http://arxiv.org/abs/2004.02088
AUTHORS:        Yang Zhao ; Chunyuan Li ; Ping Yu ; Jianfeng Gao ; Changyou Chen
COMMENTS:       The first two authors contributed equally to this manuscript. Code: https://github.com/YangNaruto/FQ-GAN
HIGHLIGHT:      In this work, we propose Feature Quantization (FQ) for the discriminator, to embed both true and fake data samples into a shared discrete space.

176, TITLE:     Discriminator Contrastive Divergence: Semi-Amortized Generative Modeling by Exploring Energy of the Discriminator
http://arxiv.org/abs/2004.01704
AUTHORS:        Yuxuan Song ; Qiwei Ye ; Minkai Xu ; Tie-Yan Liu
COMMENTS:       17 pages, 9 figures, pre-submmited to cvpr2019
HIGHLIGHT:      In this paper, we introduce the Discriminator Contrastive Divergence, which is well motivated by the property of WGAN's discriminator and the relationship between WGAN and energy-based model.

177, TITLE:     Locality Sensitive Hashing-based Sequence Alignment Using Deep Bidirectional LSTM Models
http://arxiv.org/abs/2004.02094
AUTHORS:        Neda Tavakoli
HIGHLIGHT:      This paper proposes to use deep bidirectional LSTM for sequence modeling as an approach to perform locality-sensitive hashing (LSH)-based sequence alignment.

178, TITLE:     On Tractable Representations of Binary Neural Networks
http://arxiv.org/abs/2004.02082
AUTHORS:        Weijia Shi ; Andy Shih ; Adnan Darwiche ; Arthur Choi
HIGHLIGHT:      We consider the compilation of a binary neural network's decision function into tractable representations such as Ordered Binary Decision Diagrams (OBDDs) and Sentential Decision Diagrams (SDDs).

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

180, TITLE:     Aggressive, Repetitive, Intentional, Visible, and Imbalanced: Refining Representations for Cyberbullying Classification
http://arxiv.org/abs/2004.01820
AUTHORS:        Caleb Ziems ; Ymir Vigfusson ; Fred Morstatter
COMMENTS:       12 pages, 5 figures, 22 tables, Accepted to the 14th International AAAI Conference on Web and Social Media, ICWSM'20
HIGHLIGHT:      In this study, we address the need for reliable data using an original annotation framework.

181, TITLE:     Privacy Shadow: Measuring Node Predictability and Privacy Over Time
http://arxiv.org/abs/2004.02047
AUTHORS:        Ivan Brugere ; Tanya y. Berger-Wolf
HIGHLIGHT:      We propose the privacy shadow for measuring how long a user remains predictive from an arbitrary time within the network.

182, TITLE:     Software Engineering For Automated Game Design
http://arxiv.org/abs/2004.01770
AUTHORS:        Michael Cook
HIGHLIGHT:      In this paper we explore the impact of software engineering decisions on the ability of an automated game design system to understand a game's codebase, generate new game code, and evaluate its work.

183, TITLE:     CondenseUNet: A Memory-Efficient Condensely-Connected Architecture for Bi-ventricular Blood Pool and Myocardium Segmentation
http://arxiv.org/abs/2004.02249
AUTHORS:        S. M. Kamrul Hasan ; Cristian A. Linte
COMMENTS:       7 pages, 3 figures
HIGHLIGHT:      In this work, we propose a novel memory-efficient Convolutional Neural Network (CNN) architecture as a modification of both CondenseNet, as well as DenseNet for ventricular blood-pool segmentation by introducing a bottleneck block and an upsampling path.

184, TITLE:     Arbitrary Scale Super-Resolution for Brain MRI Images
http://arxiv.org/abs/2004.02086
AUTHORS:        Chuan Tan ; Jin Zhu ; Pietro Lio'
COMMENTS:       12 pages, 8 figures, 1 table, to appear as a full paper with oral contribution in AIAI 2020
HIGHLIGHT:      A recent paper has proposed a novel meta-learning technique that uses a Weight Prediction Network to enable Super-Resolution on arbitrary scale factors using only a single neural network.

185, TITLE:     Game of Learning Bloch Equation Simulations for MR Fingerprinting
http://arxiv.org/abs/2004.02270
AUTHORS:        Mingrui Yang ; Yun Jiang ; Dan Ma ; Bhairav B. Mehta ; Mark A. Griswold
HIGHLIGHT:      Purpose: This work proposes a novel approach to efficiently generate MR fingerprints for MR fingerprinting (MRF) problems based on the unsupervised deep learning model generative adversarial networks (GAN).

186, TITLE:     Finding Covid-19 from Chest X-rays using Deep Learning on a Small Dataset
http://arxiv.org/abs/2004.02060
AUTHORS:        Lawrence O. Hall ; Rahul Paul ; Dmitry B. Goldgof ; Gregory M. Goldgof
COMMENTS:       6 pages
HIGHLIGHT:      This paper explores how useful chest X-ray images can be in diagnosing COVID-19 disease.

187, TITLE:     DeepFLASH: An Efficient Network for Learning-based Medical Image Registration
http://arxiv.org/abs/2004.02097
AUTHORS:        Jian Wang ; Miaomiao Zhang
HIGHLIGHT:      This paper presents DeepFLASH, a novel network with efficient training and inference for learning-based medical image registration.

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

189, TITLE:     AR: Auto-Repair the Synthetic Data for Neural Machine Translation
http://arxiv.org/abs/2004.02196
AUTHORS:        Shanbo Cheng ; Shaohui Kuang ; Rongxiang Weng ; Heng Yu ; Changfeng Zhu ; Weihua Luo
HIGHLIGHT:      In this paper, we propose a novel Auto- Repair (AR) framework to improve the quality of synthetic data.

190, TITLE:     FastBERT: a Self-distilling BERT with Adaptive Inference Time
http://arxiv.org/abs/2004.02178
AUTHORS:        Weijie Liu ; Peng Zhou ; Zhe Zhao ; Zhiruo Wang ; Haotang Deng ; Qi Ju
COMMENTS:       This manuscript has been accepted to appear at ACL 2020
HIGHLIGHT:      To improve their efficiency with an assured model performance, we propose a novel speed-tunable FastBERT with adaptive inference time.

191, TITLE:     Detecting and Understanding Generalization Barriers for Neural Machine Translation
http://arxiv.org/abs/2004.02181
AUTHORS:        Guanlin Li ; Lemao Liu ; Conghui Zhu ; Tiejun Zhao ; Shuming Shi
COMMENTS:       Preprint
HIGHLIGHT:      Based on the modified one, we propose three simple methods for barrier detection by the search-aware risk estimation through counterfactual generation.

192, TITLE:     Understanding Learning Dynamics for Neural Machine Translation
http://arxiv.org/abs/2004.02199
AUTHORS:        Conghui Zhu ; Guanlin Li ; Lemao Liu ; Tiejun Zhao ; Shuming Shi
COMMENTS:       Preprint
HIGHLIGHT:      In this paper we propose to understand learning dynamics of NMT by using a recent proposed technique named Loss Change Allocation (LCA)~\citep{lan-2019-loss-change-allocation}.

193, TITLE:     Arabic Offensive Language on Twitter: Analysis and Experiments
http://arxiv.org/abs/2004.02192
AUTHORS:        Hamdy Mubarak ; Ammar Rashed ; Kareem Darwish ; Younes Samih ; Ahmed Abdelali
COMMENTS:       10 pages, 6 figures, 3 tables
HIGHLIGHT:      In this paper, we focus on building effective Arabic offensive tweet detection.

194, TITLE:     The P-DESTRE: A Fully Annotated Dataset for Pedestrian Detection, Tracking, Re-Identification and Search from Aerial Devices
http://arxiv.org/abs/2004.02782
AUTHORS:        S. V. Aruna Kumar ; Ehsan Yaghoubi ; Abhijit Das ; B. S. Harish ; Hugo Proença
COMMENTS:       11 pages, 12 figures, 7 tables
HIGHLIGHT:      Accordingly, the main contributions of this paper are two-fold: 1) we announce the UAV-based P-DESTRE dataset, which is the first of its kind to provide consistent ID annotations across multiple days, making it suitable for the extremely challenging problem of person search, i.e., where no clothing information can be reliably used.

195, TITLE:     LaNet: Real-time Lane Identification by Learning Road SurfaceCharacteristics from Accelerometer Data
http://arxiv.org/abs/2004.02822
AUTHORS:        Madhumitha Harishankar ; Jun Han ; Sai Vineeth Kalluru Srinivas ; Faisal Alqarni ; Shi Su ; Shijia Pan ; Hae Young Noh ; Pei Zhang ; Marco Gruteser ; Patrick Tague
HIGHLIGHT:      In this work, we develop a deep LSTM neural network model LaNet that determines the lane vehicles are on by periodically classifying accelerometer samples collected by vehicles as they drive in real time.

196, TITLE:     Self-Supervised Scene De-occlusion
http://arxiv.org/abs/2004.02788
AUTHORS:        Xiaohang Zhan ; Xingang Pan ; Bo Dai ; Ziwei Liu ; Dahua Lin ; Chen Change Loy
COMMENTS:       Accepted to CVPR 2020 as oral. Project page: https://xiaohangzhan.github.io/projects/deocclusion/
HIGHLIGHT:      In this paper, we investigate the problem of scene de-occlusion, which aims to recover the underlying occlusion ordering and complete the invisible parts of occluded objects.

197, TITLE:     There and Back Again: Revisiting Backpropagation Saliency Methods
http://arxiv.org/abs/2004.02866
AUTHORS:        Sylvestre-Alvise Rebuffi ; Ruth Fong ; Xu Ji ; Andrea Vedaldi
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this work, we conduct a thorough analysis of backpropagation-based saliency methods and propose a single framework under which several such methods can be unified.

198, TITLE:     Rethinking Spatially-Adaptive Normalization
http://arxiv.org/abs/2004.02867
AUTHORS:        Zhentao Tan ; Dongdong Chen ; Qi Chu ; Menglei Chai ; Jing Liao ; Mingming He ; Lu Yuan ; Nenghai Yu
HIGHLIGHT:      In this paper, from a return-on-investment point of view, we present a deep analysis of the effectiveness of SPADE and observe that its advantages actually come mainly from its semantic-awareness rather than the spatial-adaptiveness.

199, TITLE:     Optical Flow Estimation in the Deep Learning Age
http://arxiv.org/abs/2004.02853
AUTHORS:        Junhwa Hur ; Stefan Roth
COMMENTS:       To appear as a book chapter in Modelling Human Motion, N. Noceti, A. Sciutti and F. Rea, Eds., Springer, 2020
HIGHLIGHT:      Alongside, we discuss some of their technical details and compare them to recapitulate which technical contribution led to the most significant accuracy improvements.

200, TITLE:     Beyond the Nav-Graph: Vision-and-Language Navigation in Continuous Environments
http://arxiv.org/abs/2004.02857
AUTHORS:        Jacob Krantz ; Erik Wijmans ; Arjun Majumdar ; Dhruv Batra ; Stefan Lee
HIGHLIGHT:      To contextualize this new task, we develop models that mirror many of the advances made in prior settings as well as single-modality baselines.

201, TITLE:     Deformable 3D Convolution for Video Super-Resolution
http://arxiv.org/abs/2004.02803
AUTHORS:        Xinyi Ying ; Longguang Wang ; Yingqian Wang ; Weidong Sheng ; Wei An ; Yulan Guo
HIGHLIGHT:      In this paper, we propose a deformable 3D convolution network (D3Dnet) to incorporate spatio-temporal information from both spatial and temporal dimensions for video SR.

202, TITLE:     FAIRS -- Soft Focus Generator and Attention for Robust Object Segmentation from Extreme Points
http://arxiv.org/abs/2004.02038
AUTHORS:        Ahmed H. Shahin ; Prateek Munjal ; Ling Shao ; Shadab Khan
HIGHLIGHT:      In this study, we present FAIRS -- a new approach to generate object segmentation from user inputs in the form of extreme points and corrective clicks.

203, TITLE:     Fine grained classification for multi-source land cover mapping
http://arxiv.org/abs/2004.01963
AUTHORS:        Yawogan Jean Eudes Gbodjo ; Dino Ienco ; Louise Leroux ; Roberto Interdonato ; Raffaelle Gaetano
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      Here, a deep learning approach is proposed to deal with multi-source land cover mapping at object level.

204, TITLE:     ObjectNet Dataset: Reanalysis and Correction
http://arxiv.org/abs/2004.02042
AUTHORS:        Ali Borji
HIGHLIGHT:      Compared with the results reported in the ObjectNet paper, we observe that around 10-15 % of the performance loss can be recovered, without any test time data augmentation.

205, TITLE:     It Is Not the Journey but the Destination: Endpoint Conditioned Trajectory Prediction
http://arxiv.org/abs/2004.02025
AUTHORS:        Karttikeya Mangalam ; Harshayu Girase ; Shreyas Agarwal ; Kuan-Hui Lee ; Ehsan Adeli ; Jitendra Malik ; Adrien Gaidon
COMMENTS:       14 pages, 6 figures, 3 tables
HIGHLIGHT:      In this work, we present Predicted Endpoint Conditioned Network (PECNet) for flexible human trajectory prediction.

206, TITLE:     Optimization of Image Embeddings for Few Shot Learning
http://arxiv.org/abs/2004.02034
AUTHORS:        Arvind Srinivasan ; Aprameya Bharadwaj ; Manasa Sathyan ; S Natarajan
COMMENTS:       6 pages, 8 figures
HIGHLIGHT:      In this paper we improve the image embeddings generated in the graph neural network solution for few shot learning.

207, TITLE:     SimAug: Learning Robust Representations from 3D Simulation for Pedestrian Trajectory Prediction in Unseen Cameras
http://arxiv.org/abs/2004.02022
AUTHORS:        Junwei Liang ; Lu Jiang ; Alexander Hauptmann
HIGHLIGHT:      We propose a method to efficiently utilize multi-view 3D simulation data for training.

208, TITLE:     Approximate Manifold Defense Against Multiple Adversarial Perturbations
http://arxiv.org/abs/2004.02183
AUTHORS:        Jay Nandy ; Wynne Hsu ; Mong Li Lee
COMMENTS:       Workshop on Machine Learning with Guarantees, NeurIPS 2019. IJCNN, 2020 (full paper)
HIGHLIGHT:      In this work, we devise an approximate manifold defense mechanism, called RBF-CNN, for image classification.

209, TITLE:     Identifying Radiological Findings Related to COVID-19 from Medical Literature
http://arxiv.org/abs/2004.01862
AUTHORS:        Yuxiao Liang ; Pengtao Xie
HIGHLIGHT:      We apply our method to the CORD-19 dataset and successfully extract a set of radiological findings that are closely tied to COVID-19.

210, TITLE:     Long-tail learning with attributes
http://arxiv.org/abs/2004.02235
AUTHORS:        Dvir Samuel ; Yuval Atzmon ; Gal Chechik
HIGHLIGHT:      Here we describe DRAGON, a novel modular architecture for long-tail learning designed to address these biases and fuse multi-modal information in face of unbalanced data. We present the first benchmark for long-tail learning with attributes and use it to evaluate DRAGON.

211, TITLE:     Large-scale spatiotemporal photonic reservoir computer for image classification
http://arxiv.org/abs/2004.02542
AUTHORS:        Piotr Antonik ; Nicolas Marsal ; Damien Rontani
COMMENTS:       12 pages, 9 figures
HIGHLIGHT:      We propose a scalable photonic architecture for implementation of feedforward and recurrent neural networks to perform the classification of handwritten digits from the MNIST database.

212, TITLE:     Bayesian optimisation of large-scale photonic reservoir computers
http://arxiv.org/abs/2004.02535
AUTHORS:        Piotr Antonik ; Nicolas Marsal ; Daniel Brunner ; Damien Rontani
COMMENTS:       10 pages, 5 figures
HIGHLIGHT:      In this work, we propose the use of Bayesian optimisation for efficient exploration of the hyper-parameter space in a minimum number of iteration.

213, TITLE:     Human action recognition with a large-scale brain-inspired photonic computer
http://arxiv.org/abs/2004.02545
AUTHORS:        Piotr Antonik ; Nicolas Marsal ; Daniel Brunner ; Damien Rontani
COMMENTS:       Authors' version before final reviews (12 pages, 4 figures)
HIGHLIGHT:      In this work, we propose a scalable photonic neuro-inspired architecture based on the reservoir computing paradigm, capable of recognising video-based human actions with state-of-the-art accuracy.

214, TITLE:     A Hierarchical Fine-Tuning Approach Based on Joint Embedding of Words and Parent Categories for Hierarchical Multi-label Text Classification
http://arxiv.org/abs/2004.02555
AUTHORS:        Yinglong Ma ; Jingpeng Zhao ; Beihong Jin
COMMENTS:       12 pages
HIGHLIGHT:      In this paper, we present a hierarchical fine-tuning deep learning approach for HMTC.

215, TITLE:     Conditions for Open-Ended Evolution in Immigration Games
http://arxiv.org/abs/2004.02720
AUTHORS:        Peter D. Turney
HIGHLIGHT:      In this paper, we call the original Immigration Game the Life Immigration Game and we call the 8,192 generalizations Immigration Games (including the Life Immigration Game).

216, TITLE:     ForecastTB -- An R Package as a Test-bench for Forecasting Methods Comparison
http://arxiv.org/abs/2004.01893
AUTHORS:        Neeraj Dhanraj Bokde ; Gorm Bruun Andersen
COMMENTS:       Preprint submitted to Energies
HIGHLIGHT:      This paper introduces the R package ForecastTB that can be used to compare the forecasting accuracy of different methods as related to the characteristics of a dataset.



==========Updates to Previous Papers==========
1, TITLE:       Generating Similarity Map for COVID-19 Transmission Dynamics with Topological Autoencoder
http://arxiv.org/abs/2004.01481
AUTHORS:        Pitoyo Hartono
HIGHLIGHT:      In this short note, the author proposes the utilization of a type of neural network to generate a global topological map for these dynamics, in which countries that share similar dynamics are mapped adjacently, while countries with significantly different dynamics are mapped far from each other.

2, TITLE:       CoKE: Contextualized Knowledge Graph Embedding
http://arxiv.org/abs/1911.02168
AUTHORS:        Quan Wang ; Pingping Huang ; Haifeng Wang ; Songtai Dai ; Wenbin Jiang ; Jing Liu ; Yajuan Lyu ; Yong Zhu ; Hua Wu
HIGHLIGHT:      This work presents Contextualized Knowledge Graph Embedding (CoKE), a novel paradigm that takes into account such contextual nature, and learns dynamic, flexible, and fully contextualized entity and relation embeddings.

3, TITLE:       Interpreting Interpretations: Organizing Attribution Methods by Criteria
http://arxiv.org/abs/2002.07985
AUTHORS:        Zifan Wang ; Piotr Mardziel ; Anupam Datta ; Matt Fredrikson
HIGHLIGHT:      In this work we expand the foundationsof human-understandable concepts with which attributionscan be interpreted beyond "importance" and its visualization; we incorporate the logical concepts of necessity andsufficiency, and the concept of proportionality.

4, TITLE:       Planning in Stochastic Environments with Goal Uncertainty
http://arxiv.org/abs/1810.08159
AUTHORS:        Sandhya Saisubramanian ; Kyle Hollins Wray ; Luis Pineda ; Shlomo Zilberstein
COMMENTS:       6 pages, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019
HIGHLIGHT:      We present the Goal Uncertain Stochastic Shortest Path (GUSSP) problem -- a general framework to model path planning and decision making in stochastic environments with goal uncertainty.

5, TITLE:       Countering Language Drift with Seeded Iterated Learning
http://arxiv.org/abs/2003.12694
AUTHORS:        Yuchen Lu ; Soumye Singhal ; Florian Strub ; Olivier Pietquin ; Aaron Courville
HIGHLIGHT:      In this paper, we propose a generic approach to counter language drift by using iterated learning.

6, TITLE:       Coherent and Archimedean choice in general Banach spaces
http://arxiv.org/abs/2002.05461
AUTHORS:        Gert de Cooman
COMMENTS:       19 pages
HIGHLIGHT:      I introduce and study a new notion of Archimedeanity for binary and non-binary choice between options that live in an abstract Banach space, through a very general class of choice models, called sets of desirable option sets.

7, TITLE:       Automatically Learning Construction Injury Precursors from Text
http://arxiv.org/abs/1907.11769
AUTHORS:        Henrietta Baker ; Matthew R. Hallowell ; Antoine J. -P. Tixier
COMMENTS:       Fixed typos, updated one figure, updated corresponding author
HIGHLIGHT:      In this study, we compare several approaches to automatically learn injury precursors from raw construction accident reports.

8, TITLE:       Pre-train and Plug-in: Flexible Conditional Text Generation with Variational Auto-Encoders
http://arxiv.org/abs/1911.03882
AUTHORS:        Yu Duan ; Canwen Xu ; Jiaxin Pei ; Jialong Han ; Chenliang Li
COMMENTS:       ACL 2020
HIGHLIGHT:      In this paper, we present a new framework named Pre-train and Plug-in Variational Auto-Encoder (PPVAE) towards flexible conditional text generation.

9, TITLE:       Evaluation of Greek Word Embeddings
http://arxiv.org/abs/1904.04032
AUTHORS:        Stamatis Outsios ; Christos Karatsalos ; Konstantinos Skianis ; Michalis Vazirgiannis
HIGHLIGHT:      This paper addresses the problem of constructing and evaluating such models for the Greek language. We created a new word analogy corpus considering the original English Word2vec word analogy corpus and some specific linguistic aspects of the Greek language as well.

10, TITLE:      A Neuro-AI Interface for Evaluating Generative Adversarial Networks
http://arxiv.org/abs/2003.03193
AUTHORS:        Zhengwei Wang ; Qi She ; Alan F. Smeaton ; Tomas E. Ward ; Graham Healy
COMMENTS:       Accepted by ICLR 2020 Workshop Bridging AI and Cognitive Science (BAICS). arXiv admin note: substantial text overlap with arXiv:1905.04243
HIGHLIGHT:      In this work, we introduce an evaluation metric called Neuroscore, for evaluating the performance of GANs, that more directly reflects psychoperceptual image quality through the utilization of brain signals.

11, TITLE:      GANprintR: Improved Fakes and Evaluation of the State-of-the-Art in Face Manipulation Detection
http://arxiv.org/abs/1911.05351
AUTHORS:        João C. Neves ; Ruben Tolosana ; Ruben Vera-Rodriguez ; Vasco Lopes ; Hugo Proença ; Julian Fierrez
HIGHLIGHT:      In this study, we focus on the synthesis of entire facial images, which is a specific type of facial manipulation.

12, TITLE:      Small-Object Detection in Remote Sensing Images with End-to-End Edge-Enhanced GAN and Object Detector Network
http://arxiv.org/abs/2003.09085
AUTHORS:        Jakaria Rabbi ; Nilanjan Ray ; Matthias Schubert ; Subir Chowdhury ; Dennis Chao
COMMENTS:       This paper contains 25 pages and submitted to MDPI remote sensing journal (under review) GitHub Repository: https://github.com/Jakaria08/EESRGAN (Implementation)
HIGHLIGHT:      We propose an architecture with three components: ESRGAN, Edge Enhancement Network (EEN), and Detection network.

13, TITLE:      CATER: A diagnostic dataset for Compositional Actions and TEmporal Reasoning
http://arxiv.org/abs/1910.04744
AUTHORS:        Rohit Girdhar ; Deva Ramanan
COMMENTS:       ICLR 2020 (oral). Webpage/code/data: https://rohitgirdhar.github.io/CATER
HIGHLIGHT:      In this work, we build a video dataset with fully observable and controllable object and scene bias, and which truly requires spatiotemporal understanding in order to be solved.

14, TITLE:      Towards Learning a Generic Agent for Vision-and-Language Navigation via Pre-training
http://arxiv.org/abs/2002.10638
AUTHORS:        Weituo Hao ; Chunyuan Li ; Xiujun Li ; Lawrence Carin ; Jianfeng Gao
COMMENTS:       To appear at CVPR 2020. The first two authors contributed equally to this manuscript. Code: https://github.com/weituo12321/PREVALENT
HIGHLIGHT:      In this paper, we present the first pre-training and fine-tuning paradigm for vision-and-language navigation (VLN) tasks.

15, TITLE:      Hierarchical Data Generator based on Tree-Structured Stick Breaking Process for Benchmarking Clustering Methods
http://arxiv.org/abs/1606.05681
AUTHORS:        Łukasz P. Olech ; Michał Spytkowski ; Halina Kwaśnicka ; Zbigniew Michalewicz
COMMENTS:       This article was submitted to Elsevier Journal
HIGHLIGHT:      In this paper we address this issue by proposing a generator of synthetic hierarchical data that can be used for benchmarking Object Cluster Hierarchy methods.

16, TITLE:      Dual-reference Age Synthesis
http://arxiv.org/abs/1908.02671
AUTHORS:        Yuan Zhou ; Bingzhang Hu ; and Jun He ; Yu Guan ; Ling Shao
HIGHLIGHT:      In this paper, we propose a novel framework taking two images as inputs, named dual-reference age synthesis (DRAS), which approaches the task differently; instead of using "hard" age information, i.e. a fixed number, our model determines the target age in a "soft" way, by employing a second reference image.

17, TITLE:      Neural Contours: Learning to Draw Lines from 3D Shapes
http://arxiv.org/abs/2003.10333
AUTHORS:        Difan Liu ; Mohamed Nabail ; Aaron Hertzmann ; Evangelos Kalogerakis
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      This paper introduces a method for learning to generate line drawings from 3D models.

18, TITLE:      Blockchain Intelligence: When Blockchain Meets Artificial Intelligence
http://arxiv.org/abs/1912.06485
AUTHORS:        Zibin Zheng ; Hong-Ning Dai ; Jiajing Wu
COMMENTS:       5 pages, 5 figures
HIGHLIGHT:      This article presents an introduction of the convergence of blockchain and artificial intelligence (namely blockchain intelligence).

19, TITLE:      Integrating Behavior Cloning and Reinforcement Learning for Improved Performance in Dense and Sparse Reward Environments
http://arxiv.org/abs/1910.04281
AUTHORS:        Vinicius G. Goecks ; Gregory M. Gremillion ; Vernon J. Lawhern ; John Valasek ; Nicholas R. Waytowich
COMMENTS:       9 pages, 5 Figures. AAMAS 2020
HIGHLIGHT:      In this work, we propose the Cycle-of-Learning (CoL) framework that uses an actor-critic architecture with a loss function that combines behavior cloning and 1-step Q-learning losses with an off-policy pre-training step from human demonstrations.

20, TITLE:      Convolutional Sparse Support Estimator Network (CSEN) From energy efficient support estimation to learning-aided Compressive Sensing
http://arxiv.org/abs/2003.00768
AUTHORS:        Mehmet Yamac ; Mete Ahishali ; Serkan Kiranyaz ; Moncef Gabbouj
HIGHLIGHT:      This study proposes a novel approach for learning such a mapping from a training set.

21, TITLE:      Online influence, offline violence: Language Use on YouTube surrounding the 'Unite the Right' rally
http://arxiv.org/abs/1908.11599
AUTHORS:        Isabelle van der Vegt ; Maximilian Mozes ; Paul Gill ; Bennett Kleinberg
COMMENTS:       pre-print (pre-peer review)
HIGHLIGHT:      To do so, we create structural topic models and measure bigram proportions in video transcripts, spanning eight weeks before to eight weeks after the rally.

22, TITLE:      A Relational Memory-based Embedding Model for Triple Classification and Search Personalization
http://arxiv.org/abs/1907.06080
AUTHORS:        Dai Quoc Nguyen ; Tu Dinh Nguyen ; Dinh Phung
COMMENTS:       To appear in Proceedings of ACL 2020
HIGHLIGHT:      To this end, we introduce a novel embedding model, named R-MeN, that explores a relational memory network to encode potential dependencies in relationship triples.

23, TITLE:      Boosting Factual Correctness of Abstractive Summarization
http://arxiv.org/abs/2003.08612
AUTHORS:        Chenguang Zhu ; William Hinthorn ; Ruochen Xu ; Qingkai Zeng ; Michael Zeng ; Xuedong Huang ; Meng Jiang
COMMENTS:       17 pages, 3 figures
HIGHLIGHT:      A commonly observed problem with abstractive summarization is the distortion or fabrication of factual information in the article.

24, TITLE:      Global Entity Disambiguation with Pretrained Contextualized Embeddings of Words and Entities
http://arxiv.org/abs/1909.00426
AUTHORS:        Ikuya Yamada ; Koki Washio ; Hiroyuki Shindo ; Yuji Matsumoto
HIGHLIGHT:      We propose a new global entity disambiguation (ED) model based on contextualized embeddings of words and entities.

25, TITLE:      Exploiting Semantics for Face Image Deblurring
http://arxiv.org/abs/2001.06822
AUTHORS:        Ziyi Shen ; Wei-Sheng Lai ; Tingfa Xu ; Jan Kautz ; Ming-Hsuan Yang
COMMENTS:       Submitted to International Journal of Computer Vision (IJCV). arXiv admin note: text overlap with arXiv:1803.03345
HIGHLIGHT:      In this paper, we propose an effective and efficient face deblurring algorithm by exploiting semantic cues via deep convolutional neural networks.

26, TITLE:      f-BRS: Rethinking Backpropagating Refinement for Interactive Segmentation
http://arxiv.org/abs/2001.10331
AUTHORS:        Konstantin Sofiiuk ; Ilia Petrov ; Olga Barinova ; Anton Konushin
HIGHLIGHT:      We propose f-BRS (feature backpropagating refinement scheme) that solves an optimization problem with respect to auxiliary variables instead of the network inputs, and requires running forward and backward pass just for a small part of a network.

27, TITLE:      Unsupervised Reinforcement Learning of Transferable Meta-Skills for Embodied Navigation
http://arxiv.org/abs/1911.07450
AUTHORS:        Juncheng Li ; Xin Wang ; Siliang Tang ; Haizhou Shi ; Fei Wu ; Yueting Zhuang ; William Yang Wang
HIGHLIGHT:      In this paper, we focus on visual navigation in the low-resource setting, where we have only a few training environments annotated with object information.

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

29, TITLE:      CNN-generated images are surprisingly easy to spot... for now
http://arxiv.org/abs/1912.11035
AUTHORS:        Sheng-Yu Wang ; Oliver Wang ; Richard Zhang ; Andrew Owens ; Alexei A. Efros
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      In this work we ask whether it is possible to create a "universal" detector for telling apart real images from these generated by a CNN, regardless of architecture or dataset used. To test this, we collect a dataset consisting of fake images generated by 11 different CNN-based image generator models, chosen to span the space of commonly used architectures today (ProGAN, StyleGAN, BigGAN, CycleGAN, StarGAN, GauGAN, DeepFakes, cascaded refinement networks, implicit maximum likelihood estimation, second-order attention super-resolution, seeing-in-the-dark).

30, TITLE:      PDANet: Pyramid Density-aware Attention Net for Accurate Crowd Counting
http://arxiv.org/abs/2001.05643
AUTHORS:        Saeed Amirgholipour ; Xiangjian He ; Wenjing Jia ; Dadong Wang ; Lei Liu
HIGHLIGHT:      In this paper, we propose a novel Pyramid Density-Aware Attention-based network, abbreviated as PDANet, that leverages the attention, pyramid scale feature and two branch decoder modules for density-aware crowd counting.

31, TITLE:      Map-Enhanced Ego-Lane Detection in the Missing Feature Scenarios
http://arxiv.org/abs/2004.01101
AUTHORS:        Xiaoliang Wang ; Yeqiang Qian ; Chunxiang Wang ; Ming Yang
COMMENTS:       There has some mistake in experiments
HIGHLIGHT:      To address this problem, previous methods have been devoted to proposing more complicated feature extraction algorithms, but they are very time-consuming and cannot deal with extreme scenarios.

32, TITLE:      KeypointNet: A Large-scale 3D Keypoint Dataset Aggregated from Numerous Human Annotations
http://arxiv.org/abs/2002.12687
AUTHORS:        Yang You ; Yujing Lou ; Chengkun Li ; Zhoujun Cheng ; Liangwei Li ; Lizhuang Ma ; Cewu Lu ; Weiming Wang
COMMENTS:       8 pages; to appear in CVPR 2020
HIGHLIGHT:      To handle the inconsistency between annotations from different people, we propose a novel method to aggregate these keypoints automatically, through minimization of a fidelity loss.

33, TITLE:      Higher-Order Function Networks for Learning Composable 3D Object Representations
http://arxiv.org/abs/1907.10388
AUTHORS:        Eric Mitchell ; Selim Engin ; Volkan Isler ; Daniel D Lee
COMMENTS:       To be published in International Conference on Learning Representations (ICLR 2020) [https://openreview.net/forum?id=HJgfDREKDB]; 19 pages
HIGHLIGHT:      We present a new approach to 3D object representation where a neural network encodes the geometry of an object directly into the weights and biases of a second 'mapping' network.

34, TITLE:      Synthetic Elastography using B-mode Ultrasound through a Deep Fully-Convolutional Neural Network
http://arxiv.org/abs/1908.03573
AUTHORS:        R. R. Wildeboer ; R. J. G. van Sloun ; C. K. Mannaerts ; P. H. Moraes ; G. Salomon ; M. C. Chammas ; H. Wijkstra ; M. Mischi
COMMENTS:       (c) 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
HIGHLIGHT:      In this work, we show how synthetic SWE (sSWE) images can be generated based on conventional B-mode imaging through deep learning.

35, TITLE:      Constrained Nonnegative Matrix Factorization for Blind Hyperspectral Unmixing incorporating Endmember Independence
http://arxiv.org/abs/2003.01041
AUTHORS:        E. M. M. B. Ekanayake ; Bhathiya Rathnayake ; G. M. R. I. Godaliyadda ; H. M. V. R. Herath ; M. P. B. Ekanayake
COMMENTS:       13 pages, 12 figures
HIGHLIGHT:      As a promising step toward finding an optimum constraint to extract endmembers, this paper presents a novel blind HU algorithm, referred to as Kurtosis-based Smooth Nonnegative Matrix Factorization (KbSNMF) which incorporates a novel constraint based on the statistical independence of the probability density functions of endmember spectra.

36, TITLE:      A Deep Learning based Pipeline for Efficient Oral Cancer Screening on Whole Slide Images
http://arxiv.org/abs/1910.10549
AUTHORS:        Jiahao Lu ; Nataša Sladoje ; Christina Runow Stark ; Eva Darai Ramqvist ; Jan-Michaél Hirsch ; Joakim Lindblad
COMMENTS:       Accepted to ICIAR 2020
HIGHLIGHT:      To facilitate large scale screening, we propose a fully automated pipeline for oral cancer detection on whole slide cytology images.

37, TITLE:      A quantum active learning algorithm for sampling against adversarial attacks
http://arxiv.org/abs/1912.03283
AUTHORS:        P. A. M. Casares ; M. A. Martin-Delgado
COMMENTS:       22 pages, 2 figures
HIGHLIGHT:      We propose a theoretical framework that allows us to think of active learning as sampling the most promising new points to be classified, so that the minimum distance between classes can be found and the theorem KH used.

38, TITLE:      Description Based Text Classification with Reinforcement Learning
http://arxiv.org/abs/2002.03067
AUTHORS:        Duo Chai ; Wei Wu ; Qinghong Han ; Fei Wu ; Jiwei Li
HIGHLIGHT:      Inspired by the current trend of formalizing NLP problems as question answering tasks, we propose a new framework for text classification, in which each category label is associated with a category description.

39, TITLE:      Index Network
http://arxiv.org/abs/1908.09895
AUTHORS:        Hao Lu ; Yutong Dai ; Chunhua Shen ; Songcen Xu
COMMENTS:       17 pages. Extended version of "Indices Matter: Learning to Index for Deep Image Matting" at arXiv:1908.00672
HIGHLIGHT:      By viewing the indices as a function of the feature map, we introduce the concept of "learning to index", and present a novel index-guided encoder-decoder framework where indices are self-learned adaptively from data and are used to guide the downsampling and upsampling stages, without extra training supervision.

40, TITLE:      MCENET: Multi-Context Encoder Network for Homogeneous Agent Trajectory Prediction in Mixed Traffic
http://arxiv.org/abs/2002.05966
AUTHORS:        Hao Cheng ; Wentong Liao ; Michael Ying Yang ; Monika Sester ; Bodo Rosenhahn
COMMENTS:       8 pages, 5 figures, code is available on https://github.com/haohao11/MCENET
HIGHLIGHT:      To this end, we propose an approach named Multi-Context Encoder Network (MCENET) that is trained by encoding both past and future scene context, interaction context and motion information to capture the patterns and variations of the future trajectories using a set of stochastic latent variables.

41, TITLE:      Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification
http://arxiv.org/abs/2003.13048
AUTHORS:        Devesh Walawalkar ; Zhiqiang Shen ; Zechun Liu ; Marios Savvides
HIGHLIGHT:      In this paper, we propose Attentive CutMix, a naturally enhanced augmentation strategy based on CutMix.

42, TITLE:      Real-Time Panoptic Segmentation from Dense Detections
http://arxiv.org/abs/1912.01202
AUTHORS:        Rui Hou ; Jie Li ; Arjun Bhargava ; Allan Raventos ; Vitor Guizilini ; Chao Fang ; Jerome Lynch ; Adrien Gaidon
COMMENTS:       12 pages, 6 figures
HIGHLIGHT:      In this paper, we propose a new single-shot panoptic segmentation network that leverages dense detections and a global self-attention mechanism to operate in real-time with performance approaching the state of the art.

43, TITLE:      EfficientDet: Scalable and Efficient Object Detection
http://arxiv.org/abs/1911.09070
AUTHORS:        Mingxing Tan ; Ruoming Pang ; Quoc V. Le
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we systematically study neural network architecture design choices for object detection and propose several key optimizations to improve efficiency.

44, TITLE:      Acceleration of Convolutional Neural Network Using FFT-Based Split Convolutions
http://arxiv.org/abs/2003.12621
AUTHORS:        Kamran Chitsaz ; Mohsen Hajabdollahi ; Nader Karimi ; Shadrokh Samavi ; Shahram Shirani
COMMENTS:       5 pages, 4 figures
HIGHLIGHT:      In this paper, a new method for CNN processing in the FFT domain is proposed, which is based on input splitting.

45, TITLE:      A concise guide to existing and emerging vehicle routing problem variants
http://arxiv.org/abs/1906.06750
AUTHORS:        Thibaut Vidal ; Gilbert Laporte ; Piotr Matl
HIGHLIGHT:      In this article, we provide a concise overview of existing and emerging problem variants.

46, TITLE:      SemanticAdv: Generating Adversarial Examples via Attribute-conditional Image Editing
http://arxiv.org/abs/1906.07927
AUTHORS:        Haonan Qiu ; Chaowei Xiao ; Lei Yang ; Xinchen Yan ; Honglak Lee ; Bo Li
HIGHLIGHT:      In this paper, we aim to explore the impact of semantic manipulation on DNNs predictions by manipulating the semantic attributes of images and generate "unrestricted adversarial examples".

47, TITLE:      A Fine-Grained Spectral Perspective on Neural Networks
http://arxiv.org/abs/1907.10599
AUTHORS:        Greg Yang ; Hadi Salman
COMMENTS:       8 pages of main text, 19 figures, 51 pages including appendix
HIGHLIGHT:      These questions seem unrelated at face value, but in this work we give all of them a common treatment from the spectral perspective.

48, TITLE:      Unsupervised Keyphrase Rubric Relationship Classification in Complex Assignments
http://arxiv.org/abs/2004.01549
AUTHORS:        Manikandan Ravikiran
COMMENTS:       v1 preprint. Working paper. More results to be added. arXiv admin note: substantial text overlap with arXiv:2003.07019
HIGHLIGHT:      In this study, we analyze both supervised and unsupervised methods to find that supervised approaches outperform unsupervised approaches and topic modelling approaches, despite data limitation with supervised approaches producing maximum results of 0.48 F1-Score and unsupervised approach producing best result of 0.31 F1-Score.

49, TITLE:      A Fine-Grained Sentiment Dataset for Norwegian
http://arxiv.org/abs/1911.12722
AUTHORS:        Lilja Øvrelid ; Petter Mæhlum ; Jeremy Barnes ; Erik Velldal
COMMENTS:       Accepted for LREC 2020
HIGHLIGHT:      We introduce NoReC_fine, a dataset for fine-grained sentiment analysis in Norwegian, annotated with respect to polar expressions, targets and holders of opinion.

50, TITLE:      Multiple Generative Models Ensemble for Knowledge-Driven Proactive Human-Computer Dialogue Agent
http://arxiv.org/abs/1907.03590
AUTHORS:        Zelin Dai ; Weitang Liu ; Guanhua Zhan
COMMENTS:       7 pages, 3 figures submitted to journal
HIGHLIGHT:      A rank-based ensemble approach was developed for boosting performance.

51, TITLE:      Directions in Abusive Language Training Data: Garbage In, Garbage Out
http://arxiv.org/abs/2004.01670
AUTHORS:        Bertie Vidgen ; Leon Derczynski
COMMENTS:       26 pages, 5 figures
HIGHLIGHT:      Data-driven analysis and detection of abusive online content covers many different tasks, phenomena, contexts, and methodologies.

52, TITLE:      Training with Streaming Annotation
http://arxiv.org/abs/2002.04165
AUTHORS:        Tongtao Zhang ; Heng Ji ; Shih-Fu Chang ; Marjorie Freedman
HIGHLIGHT:      In this paper, we address a practical scenario where training data is released in a sequence of small-scale batches and annotation in earlier phases has lower quality than the later counterparts.

53, TITLE:      Fast Learning-based Registration of Sparse 3D Clinical Images
http://arxiv.org/abs/1812.06932
AUTHORS:        Kathleen M. Lewis ; Natalia S. Rost ; John Guttag ; Adrian V. Dalca
COMMENTS:       This version was accepted to CHIL. It builds on the previous version of the paper and includes more experimental results
HIGHLIGHT:      We introduce SparseVM, a method that registers clinical-quality 3D MR scans both faster and more accurately than previously possible.

54, TITLE:      Single Image Reflection Removal through Cascaded Refinement
http://arxiv.org/abs/1911.06634
AUTHORS:        Chao Li ; Yixiao Yang ; Kun He ; Stephen Lin ; John E. Hopcroft
COMMENTS:       10 pages, 8 figures
HIGHLIGHT:      Inspired by iterative structure reduction for hidden community detection in social networks, we propose an Iterative Boost Convolutional LSTM Network (IBCLN) that enables cascaded prediction for reflection removal. Besides, we create a dataset of real-world images with reflection and ground-truth transmission layers to mitigate the problem of insufficient data.

55, TITLE:      Learning beamforming in ultrasound imaging
http://arxiv.org/abs/1812.08043
AUTHORS:        Sanketh Vedula ; Ortal Senouf ; Grigoriy Zurakhov ; Alex Bronstein ; Oleg Michailovich ; Michael Zibulevsky
HIGHLIGHT:      In this paper, we demonstrate that replacing the traditional ultrasound processing pipeline with a data-driven, learnable counterpart leads to significant improvement in image quality.

56, TITLE:      SPA-GAN: Spatial Attention GAN for Image-to-Image Translation
http://arxiv.org/abs/1908.06616
AUTHORS:        Hajar Emami ; Majid Moradi Aliabadi ; Ming Dong ; Ratna Babu Chinnam
HIGHLIGHT:      In this paper, we introduce the attention mechanism directly to the generative adversarial network (GAN) architecture and propose a novel spatial attention GAN model (SPA-GAN) for image-to-image translation tasks.

57, TITLE:      Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivation
http://arxiv.org/abs/1902.04760
AUTHORS:        Greg Yang
COMMENTS:       tldr: A theoretical tool for understanding the behavior of large width randomly initialized neural networks for almost all deep learning architectures. For a gentler introduction to Gaussian Process results here and several extensions, we recommend the reader to look at arXiv:1910.12478
HIGHLIGHT:      We unify these results by introducing a notion of a straightline \emph{tensor program} that can express most neural network computations, and we characterize its scaling limit when its tensors are large and randomized.

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

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

60, TITLE:      Stochastic first-order methods: non-asymptotic and computer-aided analyses via potential functions
http://arxiv.org/abs/1902.00947
AUTHORS:        Adrien Taylor ; Francis Bach
COMMENTS:       Accepted to COLT2019; 12 pages + appendix; code available at https://github.com/AdrienTaylor/Potential-functions-for-first-order-methods. [V5: typos & minor improvements to appendix E.4]
HIGHLIGHT:      We provide a novel computer-assisted technique for systematically analyzing first-order methods for optimization.

61, TITLE:      Software Language Comprehension using a Program-Derived Semantic Graph
http://arxiv.org/abs/2004.00768
AUTHORS:        Roshni G. Iyer ; Yizhou Sun ; Wei Wang ; Justin Gottschlich
HIGHLIGHT:      In this paper, we describe the PSG and its fundamental structural differences to the Aroma's SPT.

62, TITLE:      Shallow Discourse Annotation for Chinese TED Talks
http://arxiv.org/abs/2003.04032
AUTHORS:        Wanqiu Long ; Xinyi Cai ; James E. M. Reid ; Bonnie Webber ; Deyi Xiong
HIGHLIGHT:      The current work contributes a new resource for Chinese Language Technology and for Chinese-English translation, in the form of a set of TED talks (some originally given in English, some in Chinese) that have been annotated with discourse relations in the style of the Penn Discourse TreeBank, adapted to properties of Chinese text that are not present in English.

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

64, TITLE:      Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
http://arxiv.org/abs/1907.11932
AUTHORS:        Di Jin ; Zhijing Jin ; Joey Tianyi Zhou ; Peter Szolovits
COMMENTS:       AAAI 2020 (Oral)
HIGHLIGHT:      In this paper, we present TextFooler, a simple but strong baseline to generate natural adversarial text.

65, TITLE:      Should All Cross-Lingual Embeddings Speak English?
http://arxiv.org/abs/1911.03058
AUTHORS:        Antonios Anastasopoulos ; Graham Neubig
COMMENTS:       accepted for publication at ACL 2020
HIGHLIGHT:      Evaluating established methods over all these language pairs sheds light into their suitability and presents new challenges for the field.

66, TITLE:      A Resource for Computational Experiments on Mapudungun
http://arxiv.org/abs/1912.01772
AUTHORS:        Mingjun Duan ; Carlos Fasola ; Sai Krishna Rallabandi ; Rodolfo M. Vega ; Antonios Anastasopoulos ; Lori Levin ; Alan W Black
COMMENTS:       accepted at LREC 2020
HIGHLIGHT:      We present a resource for computational experiments on Mapudungun, a polysynthetic indigenous language spoken in Chile with upwards of 200 thousand speakers.

67, TITLE:      How Can BERT Help Lexical Semantics Tasks?
http://arxiv.org/abs/1911.02929
AUTHORS:        Yile Wang ; Leyang Cui ; Yue Zhang
HIGHLIGHT:      We address this issue by making use of dynamic embeddings as word representations in training static embeddings, thereby leveraging their strong representation power for disambiguating context information.

68, TITLE:      Neural Voxel Renderer: Learning an Accurate and Controllable Rendering Tool
http://arxiv.org/abs/1912.04591
AUTHORS:        Konstantinos Rematas ; Vittorio Ferrari
COMMENTS:       Additional results: http://www.krematas.com/nvr/index.html
HIGHLIGHT:      We present a neural rendering framework that maps a voxelized scene into a high quality image.

69, TITLE:      The perceptual boost of visual attention is task-dependent in naturalistic settings
http://arxiv.org/abs/2003.00882
AUTHORS:        Freddie Bickford Smith ; Xiaoliang Luo ; Brett D. Roads ; Bradley C. Love
COMMENTS:       Published as a workshop paper at "Bridging AI and Cognitive Science" (ICLR 2020)
HIGHLIGHT:      We aim to answer this with a large-scale computational experiment.

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

71, TITLE:      Real-Time Fruit Recognition and Grasping Estimation for Autonomous Apple Harvesting
http://arxiv.org/abs/2003.13298
AUTHORS:        Hanwen Kang ; Chao Chen
HIGHLIGHT:      In this research, a fully neural network based visual perception framework for autonomous apple harvesting is proposed.

72, TITLE:      C-DLinkNet: considering multi-level semantic features for human parsing
http://arxiv.org/abs/2001.11690
AUTHORS:        Yu Lu ; Muyan Feng ; Ming Wu ; Chuang Zhang
HIGHLIGHT:      In this work, we proposed an end-to-end model called C-DLinkNet based on LinkNet, which contains a new module named Smooth Module to combine the multi-level features in Decoder part.

73, TITLE:      Binary Classification of Alzheimer Disease using sMRI Imaging modality and Deep Learning
http://arxiv.org/abs/1809.06209
AUTHORS:        Ahsan Bin Tufail ; Qiu-Na Zhang ; Yong-Kui Ma
HIGHLIGHT:      Different from these methods, this paper proposes to construct multiple deep 2D convolutional neural networks (2D-CNNs) to learn the various features from local brain images which are combined to make the final classification for AD diagnosis.

74, TITLE:      Robustness Guarantees for Deep Neural Networks on Videos
http://arxiv.org/abs/1907.00098
AUTHORS:        Min Wu ; Marta Kwiatkowska
COMMENTS:       To appear in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
HIGHLIGHT:      In this paper, we consider the robustness of deep neural networks on videos, which comprise both the spatial features of individual frames extracted by a convolutional neural network and the temporal dynamics between adjacent frames captured by a recurrent neural network.

75, TITLE:      Explain Your Move: Understanding Agent Actions Using Specific and Relevant Feature Attribution
http://arxiv.org/abs/1912.12191
AUTHORS:        Nikaash Puri ; Sukriti Verma ; Piyush Gupta ; Dhruv Kayastha ; Shripad Deshmukh ; Balaji Krishnamurthy ; Sameer Singh
COMMENTS:       Accepted at the International Conference on Learning Representations (ICLR) 2020
HIGHLIGHT:      We compare SARFA with existing approaches on agents trained to play board games (Chess and Go) and Atari games (Breakout, Pong and Space Invaders).

76, TITLE:      Generative Hybrid Representations for Activity Forecasting with No-Regret Learning
http://arxiv.org/abs/1904.06250
AUTHORS:        Jiaqi Guan ; Ye Yuan ; Kris M. Kitani ; Nicholas Rhinehart
COMMENTS:       Oral presentation at CVPR 2020
HIGHLIGHT:      In this work, we develop an efficient deep generative model to jointly forecast a person's future discrete actions and continuous motions.

77, TITLE:      Extreme Relative Pose Network under Hybrid Representations
http://arxiv.org/abs/1912.11695
AUTHORS:        Zhenpei Yang ; Siming Yan ; Qixing Huang
COMMENTS:       Accepted in CVPR 2020
HIGHLIGHT:      In this paper, we introduce a novel RGB-D based relative pose estimation approach that is suitable for small-overlapping or non-overlapping scans and can output multiple relative poses.

78, TITLE:      Frontal Low-rank Random Tensors for Fine-grained Action Segmentation
http://arxiv.org/abs/1906.01004
AUTHORS:        Yan Zhang ; Krikamol Muandet ; Qianli Ma ; Heiko Neumann ; Siyu Tang
COMMENTS:       19 pages (4 pages appendix), 3 figures. Revised theories and models, new experiments
HIGHLIGHT:      In this paper, we propose an approach to representing high-order information for temporal action segmentation via a simple yet effective bilinear form.

79, TITLE:      Polynomial Invariant Generation for Non-deterministic Recursive Programs
http://arxiv.org/abs/1902.04373
AUTHORS:        Krishnendu Chatterjee ; Hongfei Fu ; Amir Kafshdar Goharshady ; Ehsan Kafshdar Goharshady
COMMENTS:       A conference version of this article appears in PLDI 2020
HIGHLIGHT:      We present a sound and semi-complete method based on positivstellensaetze, i.e. theorems in semi-algebraic geometry that characterize positive polynomials over a semi-algebraic set.

80, TITLE:      IntrA: 3D Intracranial Aneurysm Dataset for Deep Learning
http://arxiv.org/abs/2003.02920
AUTHORS:        Xi Yang ; Ding Xia ; Taichi Kin ; Takeo Igarashi
COMMENTS:       Accepted by cvpr2020, camera-ready version will be uploaded later
HIGHLIGHT:      In this paper, instead of 2D medical images, we introduce an open-access 3D intracranial aneurysm dataset, IntrA, that makes the application of points-based and mesh-based classification and segmentation models available. We provide a large-scale benchmark of classification and part segmentation by testing state-of-the-art networks.

81, TITLE:      AQPDCITY Dataset: Picture-Based PM Monitoring in the Urban Area of Big Cities
http://arxiv.org/abs/2003.09784
AUTHORS:        Yonghui Zhang ; Ke Gu
HIGHLIGHT:      It is highly desired to devise a method that can obtain the PM concentration at any location for the following air quality control in time.

82, TITLE:      A computer-aided diagnosis system using artificial intelligence for hip fractures significantly improves the diagnostic rate of residents. -Multi-institutional joint Development Research
http://arxiv.org/abs/2003.12443
AUTHORS:        Yoichi Sato ; Yasuhiko Takegami ; Takamune Asamoto ; Yutaro Ono ; Ryosuke Goto ; Asahi Kitamura ; Seiwa Honda
COMMENTS:       6 pages, 3 tables, 7 figures. / author's homepage : https://www.fracture-ai.org
HIGHLIGHT:      We used the EffectiventNet-B4 framework with Pytorch 1.3 and this http URL 1.0.

83, TITLE:      Cumulant-free closed-form formulas for some common (dis)similarities between densities of an exponential family
http://arxiv.org/abs/2003.02469
AUTHORS:        Frank Nielsen ; Richard Nock
COMMENTS:       32 pages
HIGHLIGHT:      In this work, we report (dis)similarity formulas which bypass the explicit use of the cumulant function and highlight the role of quasi-arithmetic means and their multivariate mean operator extensions.

84, TITLE:      MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
http://arxiv.org/abs/2002.10957
AUTHORS:        Wenhui Wang ; Furu Wei ; Li Dong ; Hangbo Bao ; Nan Yang ; Ming Zhou
COMMENTS:       Code and models: https://github.com/microsoft/unilm/tree/master/minilm
HIGHLIGHT:      In this work, we present a simple and effective approach to compress large Transformer (Vaswani et al., 2017) based pre-trained models, termed as deep self-attention distillation.

85, TITLE:      Enhancing Machine Translation with Dependency-Aware Self-Attention
http://arxiv.org/abs/1909.03149
AUTHORS:        Emanuele Bugliarello ; Naoaki Okazaki
COMMENTS:       Update title and affiliations, and fix typos
HIGHLIGHT:      In this work, we explore different approaches to make such models syntactically aware.

86, TITLE:      LSTM-based Whisper Detection
http://arxiv.org/abs/1809.07832
AUTHORS:        Zeynab Raeesy ; Kellen Gillespie ; Zhenpei Yang ; Chengyuan Ma ; Thomas Drugman ; Jiacheng Gu ; Roland Maas ; Ariya Rastrow ; Björn Hoffmeister
HIGHLIGHT:      This article presents a whisper speech detector in the far-field domain.

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

88, TITLE:      Grounded Conversation Generation as Guided Traverses in Commonsense Knowledge Graphs
http://arxiv.org/abs/1911.02707
AUTHORS:        Houyu Zhang ; Zhenghao Liu ; Chenyan Xiong ; Zhiyuan Liu
HIGHLIGHT:      This paper presents a new conversation generation model, ConceptFlow, which leverages commonsense knowledge graphs to explicitly model conversation flows.

89, TITLE:      On the distribution of runners on a circle
http://arxiv.org/abs/1906.02511
AUTHORS:        Pavel Hrubes
HIGHLIGHT:      We show that, at some time, there will exist a sector $S$ which contains at least $|S|n+ \Omega(\sqrt{k})$ runners.

90, TITLE:      Recognizing License Plates in Real-Time
http://arxiv.org/abs/1906.04376
AUTHORS:        Xuewen Yang ; Xin Wang
COMMENTS:       License Plate Detection and Recognition, Computer Vision, Supervised Learning
HIGHLIGHT:      To enable real-time and accurate license plate recognition, in this work, we propose a set of techniques: 1) a contour reconstruction method along with edge-detection to quickly detect the candidate plates; 2) a simple zero-one-alternation scheme to effectively remove the fake top and bottom borders around plates to facilitate more accurate segmentation of characters on plates; 3) a set of techniques to augment the training data, incorporate SIFT features into the CNN network, and exploit transfer learning to obtain the initial parameters for more effective training; and 4) a two-phase verification procedure to determine the correct plate at low cost, a statistical filtering in the plate detection stage to quickly remove unwanted candidates, and the accurate CR results after the CR process to perform further plate verification without additional processing.

91, TITLE:      Learning from Synthetic Animals
http://arxiv.org/abs/1912.08265
AUTHORS:        Jiteng Mu ; Weichao Qiu ; Gregory Hager ; Alan Yuille
HIGHLIGHT:      In this paper, we use synthetic images and ground truth generated from CAD animal models to address this challenge.

92, TITLE:      Neural Point-Based Graphics
http://arxiv.org/abs/1906.08240
AUTHORS:        Kara-Ali Aliev ; Artem Sevastopolsky ; Maria Kolos ; Dmitry Ulyanov ; Victor Lempitsky
HIGHLIGHT:      We present a new point-based approach for modeling the appearance of real scenes.

93, TITLE:      CLAREL: Classification via retrieval loss for zero-shot learning
http://arxiv.org/abs/1906.11892
AUTHORS:        Boris N. Oreshkin ; Negar Rostamzadeh ; Pedro O. Pinheiro ; Christopher Pal
HIGHLIGHT:      We propose an instance-based deep metric learning approach in joint visual and textual space.

94, TITLE:      Sperm Detection and Tracking in Phase-Contrast Microscopy Image Sequences using Deep Learning and Modified CSR-DCF
http://arxiv.org/abs/2002.04034
AUTHORS:        Mohammad reza Mohammadi ; Mohammad Rahimzadeh ; Abolfazl Attar
HIGHLIGHT:      In this article, we used RetinaNet, a deep fully convolutional neural network as the object detector.

95, TITLE:      Minimal Solutions for Relative Pose with a Single Affine Correspondence
http://arxiv.org/abs/1912.10776
AUTHORS:        Banglei Guan ; Ji Zhao ; Zhang Li ; Fang Sun ; Friedrich Fraundorfer
COMMENTS:       IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020
HIGHLIGHT:      In this paper we present four cases of minimal solutions for two-view relative pose estimation by exploiting the affine transformation between feature points and we demonstrate efficient solvers for these cases.

96, TITLE:      DCA: Diversified Co-Attention towards Informative Live Video Commenting
http://arxiv.org/abs/1911.02739
AUTHORS:        Zhihan Zhang ; Zhiyi Yin ; Shuhuai Ren ; Xinhang Li ; Shicheng Li
HIGHLIGHT:      In this paper, we aim to collect diversified information from video and text for informative comment generation.

97, TITLE:      PanDA: Panoptic Data Augmentation
http://arxiv.org/abs/1911.12317
AUTHORS:        Yang Liu ; Pietro Perona ; Markus Meister
HIGHLIGHT:      In this paper we present an efficient and novel panoptic data augmentation (PanDA) method which operates exclusively in pixel space, requires no additional data or training, and is computationally cheap to implement.

98, TITLE:      Self-supervised Learning for ECG-based Emotion Recognition
http://arxiv.org/abs/1910.07497
AUTHORS:        Pritam Sarkar ; Ali Etemad
COMMENTS:       Accepted, 45th IEEE International Conference on Acoustics, Speech, and Signal Processing
HIGHLIGHT:      We present an electrocardiogram (ECG) -based emotion recognition system using self-supervised learning.

99, TITLE:      A Benchmark for Point Clouds Registration Algorithms
http://arxiv.org/abs/2003.12841
AUTHORS:        Simone Fontana ; Daniele Cattaneo ; Augusto Luis Ballardini ; Matteo Vaghi ; Domenico Giorgio Sorrenti
HIGHLIGHT:      This work aims at encouraging authors to use a public and shared benchmark, instead than data collected ad-hoc, to ensure objectivity and repeatability, two fundamental characteristics in any scientific field. For these reasons, we developed a benchmark, for localization and mapping applications, using multiple publicly available datasets. Along with the data, we provide a broad set of registration problems, chosen to cover different types of initial misalignment, various degrees of overlap, and different kinds of registration problems.

100, TITLE:     The Prolog Debugger and Declarative Programming. Examples
http://arxiv.org/abs/2003.01422
AUTHORS:        Włodzimierz Drabent
COMMENTS:       11 pages, 8 figures (an example added to the previous version + a few corrections)
HIGHLIGHT:      This paper contains examples for a companion paper "The Prolog Debugger and Declarative Programming", which discusses (in)adequacy of the Prolog debugger for declarative programming.

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

102, TITLE:     Joint learning of cartesian undersampling and reconstruction for accelerated MRI
http://arxiv.org/abs/1905.09324
AUTHORS:        Tomer Weiss ; Sanketh Vedula ; Ortal Senouf ; Oleg Michailovich ; Michael Zibulevsky ; Alex Bronstein
COMMENTS:       ICASSP 2020
HIGHLIGHT:      To this end, we propose an algorithm for training the combined acquisition-reconstruction pipeline end-to-end in a differentiable way.

103, TITLE:     On the design of convolutional neural networks for automatic detection of Alzheimer's disease
http://arxiv.org/abs/1911.03740
AUTHORS:        Sheng Liu ; Chhavi Yadav ; Carlos Fernandez-Granda ; Narges Razavian
COMMENTS:       Machine Learning for Health Workshop, NeurIPS2019. Authors Fernandez-Granda and Razavian are joint last authors
HIGHLIGHT:      Early detection is a crucial goal in the study of Alzheimer's Disease (AD).

104, TITLE:     A Novel Hierarchical Binary Tagging Framework for Relational Triple Extraction
http://arxiv.org/abs/1909.03227
AUTHORS:        Zhepei Wei ; Jianlin Su ; Yue Wang ; Yuan Tian ; Yi Chang
COMMENTS:       To appear in ACL 2020. The camera ready paper is still in preparation
HIGHLIGHT:      In this work, we introduce a fresh perspective to revisit the relational triple extraction task and propose a novel Hierarchical Binary Tagging (HBT) framework derived from a principled problem formulation.

105, TITLE:     CodeBERT: A Pre-Trained Model for Programming and Natural Languages
http://arxiv.org/abs/2002.08155
AUTHORS:        Zhangyin Feng ; Daya Guo ; Duyu Tang ; Nan Duan ; Xiaocheng Feng ; Ming Gong ; Linjun Shou ; Bing Qin ; Ting Liu ; Daxin Jiang ; Ming Zhou
COMMENTS:       10 pages
HIGHLIGHT:      We present CodeBERT, a bimodal pre-trained model for programming language (PL) and nat-ural language (NL). Furthermore, to investigate what type of knowledge is learned in CodeBERT, we construct a dataset for NL-PL probing, and evaluate in a zero-shot setting where parameters of pre-trained models are fixed.

106, TITLE:     Predicting Unplanned Readmissions with Highly Unstructured Data
http://arxiv.org/abs/2003.11622
AUTHORS:        Constanza Fierro ; Jorge Pérez ; Javier Mora
HIGHLIGHT:      In this paper we propose a deep learning architecture for predicting unplanned readmissions that consumes data that is significantly less structured compared with previous models in the literature.

107, TITLE:     Bridging the gap between AI and Healthcare sides: towards developing clinically relevant AI-powered diagnosis systems
http://arxiv.org/abs/2001.03923
AUTHORS:        Changhee Han ; Leonardo Rundo ; Kohei Murao ; Takafumi Nemoto ; Hideki Nakayama
COMMENTS:       13 pages, 2 figure, accepted to AIAI 2020
HIGHLIGHT:      The workshop reveals the intrinsic gap between AI/Healthcare sides and solutions on Why (i.e., clinical significance/interpretation) and How (i.e., data acquisition, commercial deployment, and safety/feeling safe).

108, TITLE:     CAGNet: Content-Aware Guidance for Salient Object Detection
http://arxiv.org/abs/1911.13168
AUTHORS:        Sina Mohammadi ; Mehrdad Noori ; Ali Bahri ; Sina Ghofrani Majelan ; Mohammad Havaei
COMMENTS:       25 pages, 10 figures, 5 tables, Accepted by Elsevier, Pattern Recognition
HIGHLIGHT:      To handle these complex scenarios, we propose a Feature Guide Network which exploits the nature of low-level and high-level features to i) make foreground and background regions more distinct and suppress the non-salient regions which have "salient-like" appearance; ii) assign foreground label to different-looking salient regions.

109, TITLE:     SampleNet: Differentiable Point Cloud Sampling
http://arxiv.org/abs/1912.03663
AUTHORS:        Itai Lang ; Asaf Manor ; Shai Avidan
COMMENTS:       CVPR 2020
HIGHLIGHT:      We introduce a novel differentiable relaxation for point cloud sampling that approximates sampled points as a mixture of points in the primary input cloud.

110, TITLE:     Category-wise Attack: Transferable Adversarial Examples for Anchor Free Object Detection
http://arxiv.org/abs/2003.04367
AUTHORS:        Quanyu Liao ; Xin Wang ; Bin Kong ; Siwei Lyu ; Youbing Yin ; Qi Song ; Xi Wu
HIGHLIGHT:      In this work, we aim to present an effective and efficient algorithm to generate adversarial examples to attack anchor-free object models based on two approaches.

111, TITLE:     Augmented Q Imitation Learning (AQIL)
http://arxiv.org/abs/2004.00993
AUTHORS:        Xiao Lei Zhang ; Anish Agarwal
COMMENTS:       5 pages
HIGHLIGHT:      This paper proposes Augmented Q-Imitation-Learning, a method by which deep reinforcement learning convergence can be accelerated by applying Q-imitation-learning as the initial training process in traditional Deep Q-learning.

112, TITLE:     Opportunities and Challenges in Deep Learning Methods on Electrocardiogram Data: A Systematic Review
http://arxiv.org/abs/2001.01550
AUTHORS:        Shenda Hong ; Yuxi Zhou ; Junyuan Shang ; Cao Xiao ; Jimeng Sun
HIGHLIGHT:      Objective: This paper conducts a systematic review of deep learning methods on ECG data from both model and application perspectives.