2020.04.09.txt

==========New Papers==========
1, TITLE:       Flexible Transmitter Network
http://arxiv.org/abs/2004.03839
AUTHORS:        Shao-Qun Zhang ; Zhi-Hua Zhou
HIGHLIGHT:      In this paper, we propose the Flexible Transmitter (FT) model, a novel bio-plausible neuron with flexible plasticity.

2, TITLE:       Attentive Normalization for Conditional Image Generation
http://arxiv.org/abs/2004.03828
AUTHORS:        Yi Wang ; Ying-Cong Chen ; Xiangyu Zhang ; Jian Sun ; Jiaya Jia
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we characterize long-range dependence with attentive normalization (AN), which is an extension to traditional instance normalization.

3, TITLE:       Exploring Versatile Generative Language Model Via Parameter-Efficient Transfer Learning
http://arxiv.org/abs/2004.03829
AUTHORS:        Zhaojiang Lin ; Andrea Madotto ; Pascale Fung
HIGHLIGHT:      In this paper, we propose an effective way to fine-tune multiple down-stream generation tasks simultaneously using a single, large pre-trained model.

4, TITLE:       SIA: A Scalable Interoperable Annotation Server for Biomedical Named Entities
http://arxiv.org/abs/2004.03822
AUTHORS:        Johannes Kirschnick ; Philippe Thomas ; Roland Roller ; Leonhard Hennig
COMMENTS:       11 pages, 2 figures, published in Journal of Cheminformatics
HIGHLIGHT:      We describe SIA (Scalable Interoperable Annotation Server), our contribution to the BeCalm-Technical interoperability and performance of annotation servers (BeCalm-TIPS) task, a scalable, extensible, and robust annotation service.

5, TITLE:       Change Detection in Heterogeneous Optical and SAR Remote Sensing Images via Deep Homogeneous Feature Fusion
http://arxiv.org/abs/2004.03830
AUTHORS:        Xiao Jiang ; Gang Li ; Yu Liu ; Xiao-Ping Zhang ; You He
COMMENTS:       15 pages, 14 figures, Accepted by IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
HIGHLIGHT:      To solve this problem, this paper presents a new homogeneous transformation model termed deep homogeneous feature fusion (DHFF) based on image style transfer (IST).

6, TITLE:       Shape-Constrained Regression using Sum of Squares Polynomials
http://arxiv.org/abs/2004.03853
AUTHORS:        Mihaela Curmei ; Georgina Hall
HIGHLIGHT:      We consider the problem of fitting a polynomial to a set of data points, each data point consisting of a feature vector and a response variable.

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

8, TITLE:       Multi-Head Attention-based Probabilistic Vehicle Trajectory Prediction
http://arxiv.org/abs/2004.03842
AUTHORS:        Hayoung Kim ; Dongchan Kim ; Gihoon Kim ; Jeongmin Cho ; Kunsoo Huh
COMMENTS:       6 pages, 5 figures, 2020 IEEE Intelligent Vehicles Symposium (IV)
HIGHLIGHT:      This paper presents online-capable deep learning model for probabilistic vehicle trajectory prediction.

9, TITLE:       Poor Man's BERT: Smaller and Faster Transformer Models
http://arxiv.org/abs/2004.03844
AUTHORS:        Hassan Sajjad ; Fahim Dalvi ; Nadir Durrani ; Preslav Nakov
COMMENTS:       12
HIGHLIGHT:      To address this issue, we explore a number of memory-light model reduction strategies that do not require model pre-training from scratch.

10, TITLE:      Structure-Level Knowledge Distillation For Multilingual Sequence Labeling
http://arxiv.org/abs/2004.03846
AUTHORS:        Xinyu Wang ; Yong Jiang ; Nguyen Bach ; Tao Wang ; Fei Huang ; Kewei Tu
COMMENTS:       Accepted to ACL 2020, submission version
HIGHLIGHT:      In this paper, we propose to reduce the gap between monolingual models and the unified multilingual model by distilling the structural knowledge of several monolingual models (teachers) to the unified multilingual model (student).

11, TITLE:      MNIST-MIX: A Multi-language Handwritten Digit Recognition Dataset
http://arxiv.org/abs/2004.03848
AUTHORS:        Weiwei Jiang
COMMENTS:       3 pages, 1 figure, 2 tables
HIGHLIGHT:      In this letter, we contribute a multi-language handwritten digit recognition dataset named MNIST-MIX, which is the largest dataset of the same type in terms of both languages and data samples.

12, TITLE:      ShanghaiTech at MRP 2019: Sequence-to-Graph Transduction with Second-Order Edge Inference for Cross-Framework Meaning Representation Parsing
http://arxiv.org/abs/2004.03849
AUTHORS:        Xinyu Wang ; Yixian Liu ; Zixia Jia ; Chengyue Jiang ; Kewei Tu
COMMENTS:       Accepted to CoNLL 2019 Shared Task
HIGHLIGHT:      This paper presents the system used in our submission to the \textit{CoNLL 2019 shared task: Cross-Framework Meaning Representation Parsing}.

13, TITLE:      Evaluating Machines by their Real-World Language Use
http://arxiv.org/abs/2004.03607
AUTHORS:        Rowan Zellers ; Ari Holtzman ; Elizabeth Clark ; Lianhui Qin ; Ali Farhadi ; Yejin Choi
COMMENTS:       Project page at https://rowanzellers.com/advice
HIGHLIGHT:      We introduce TuringAdvice, a new challenge for language understanding systems.

14, TITLE:      Diverse, Controllable, and Keyphrase-Aware: A Corpus and Method for News Multi-Headline Generation
http://arxiv.org/abs/2004.03875
AUTHORS:        Dayiheng Liu ; Yeyun Gong ; Jie Fu ; Wei Liu ; Yu Yan ; Bo Shao ; Daxin Jiang ; Jiancheng Lv ; Nan Duan
COMMENTS:       11 pages, 3 figures
HIGHLIGHT:      In this paper, we propose generating multiple headlines with keyphrases of user interests, whose main idea is to generate multiple keyphrases of interest to users for the news first, and then generate multiple keyphrase-relevant headlines.

15, TITLE:      Efficient long-distance relation extraction with DG-SpanBERT
http://arxiv.org/abs/2004.03636
AUTHORS:        Jun Chen ; Robert Hoehndorf ; Mohamed Elhoseiny ; Xiangliang Zhang
HIGHLIGHT:      Here, we propose a novel SpanBERT-based graph convolutional network (DG-SpanBERT) that extracts semantic features from a raw sentence using the pre-trained language model SpanBERT and a graph convolutional network to pool latent features.

16, TITLE:      Monte-Carlo Siamese Policy on Actor for Satellite Image Super Resolution
http://arxiv.org/abs/2004.03879
AUTHORS:        Litu Rout ; Saumyaa Shah ; S Manthira Moorthi ; Debajyoti Dhar
COMMENTS:       Computer Vision and Pattern Recognition (CVPR) Workshop on Large Scale Computer Vision for Remote Sensing Imagery
HIGHLIGHT:      In this study, we explore the plausible usage of RL in super resolution of remote sensing imagery.

17, TITLE:      CNN in CT Image Segmentation: Beyound Loss Function for Expoliting Ground Truth Images
http://arxiv.org/abs/2004.03882
AUTHORS:        Youyi Song ; Zhen Yu ; Teng Zhou ; Jeremy Yuen-Chun Teoh ; Baiying Lei ; Kup-Sze Choi ; Jing Qin
COMMENTS:       4 pages, 3 figures, and having been accepted by ISBI 2020
HIGHLIGHT:      We here present a novel and practical method that exploits GT images beyond the loss function.

18, TITLE:      Re-translation versus Streaming for Simultaneous Translation
http://arxiv.org/abs/2004.03643
AUTHORS:        Naveen Arivazhagan ; Colin Cherry ; Wolfgang Macherey ; George Foster
HIGHLIGHT:      We study a related problem in which revisions to the hypothesis beyond strictly appending words are permitted.

19, TITLE:      PatchVAE: Learning Local Latent Codes for Recognition
http://arxiv.org/abs/2004.03623
AUTHORS:        Kamal Gupta ; Saurabh Singh ; Abhinav Shrivastava
COMMENTS:       To appear at CVPR 2020
HIGHLIGHT:      Drawing inspiration from the mid-level representation discovery work, we propose PatchVAE, that reasons about images at patch level.

20, TITLE:      The relationship between Fully Connected Layers and number of classes for the analysis of retinal images
http://arxiv.org/abs/2004.03624
AUTHORS:        Ajna Ram ; d Constantino Carlos Reyes-Aldasoro
HIGHLIGHT:      This paper hence aims to find the relationship between number of classes and number of fully-connected layers.

21, TITLE:      S2A: Wasserstein GAN with Spatio-Spectral Laplacian Attention for Multi-Spectral Band Synthesis
http://arxiv.org/abs/2004.03867
AUTHORS:        Litu Rout ; Indranil Misra ; S Manthira Moorthi ; Debajyoti Dhar
COMMENTS:       Computer Vision and Pattern Recognition (CVPR) Workshop on Large Scale Computer Vision for Remote Sensing Imagery
HIGHLIGHT:      In this study, we intend to address synthesis of high resolution multi-spectral satellite imagery using adversarial learning.

22, TITLE:      Internal and external pressures on language emergence: least effort, object constancy and frequency
http://arxiv.org/abs/2004.03868
AUTHORS:        Diana Rodríguez Luna ; Edoardo Maria Ponti ; Dieuwke Hupkes ; Elia Bruni
HIGHLIGHT:      In this paper, we propose some realistic sources of pressure on communication that avert this outcome.

23, TITLE:      TypeNet: Scaling up Keystroke Biometrics
http://arxiv.org/abs/2004.03627
AUTHORS:        Alejandro Acien ; John V. Monaco ; Aythami Morales ; Ruben Vera-Rodriguez ; Julian Fierrez
HIGHLIGHT:      We study the suitability of keystroke dynamics to authenticate 100K users typing free-text.

24, TITLE:      Training Neural Networks to Produce Compatible Features
http://arxiv.org/abs/2004.03898
AUTHORS:        Michael Gygli ; Jasper Uijlings ; Vittorio Ferrari
HIGHLIGHT:      We propose and compare several different approaches to accomplish compatibility.

25, TITLE:      Guessing What's Plausible But Remembering What's True: Accurate Neural Reasoning for Question-Answering
http://arxiv.org/abs/2004.03658
AUTHORS:        Haitian Sun ; Andrew O. Arnold ; Tania Bedrax-Weiss ; Fernando Pereira ; William W. Cohen
HIGHLIGHT:      Our approach introduces two new mechanisms for KB reasoning: neural retrieval over a set of embedded triples, and "memorization" of highly specific information with a compact sketch structure.

26, TITLE:      The Russian Drug Reaction Corpus and Neural Models for Drug Reactions and Effectiveness Detection in User Reviews
http://arxiv.org/abs/2004.03659
AUTHORS:        Elena Tutubalina ; Ilseyar Alimova ; Zulfat Miftahutdinov ; Andrey Sakhovskiy ; Valentin Malykh ; Sergey Nikolenko
COMMENTS:       9 pages, 9 tables, 4 figures
HIGHLIGHT:      For the sentence classification task, our model achieves the macro F1 score of 68.82% gaining 7.47% over the score of BERT model trained on Russian data.

27, TITLE:      Query-controllable Video Summarization
http://arxiv.org/abs/2004.03661
AUTHORS:        Jia-Hong Huang ; Marcel Worring
COMMENTS:       This paper is accepted by ACM International Conference on Multimedia Retrieval (ICMR), 2020
HIGHLIGHT:      In this work, we introduce a method which takes a text-based query as input and generates a video summary corresponding to it. To foster the research of query-controllable video summarization and conduct our experiments, we introduce a dataset that contains frame-based relevance score labels.

28, TITLE:      Causal Relational Learning
http://arxiv.org/abs/2004.03644
AUTHORS:        Babak Salimi ; Harsh Parikh ; Moe Kayali ; Sudeepa Roy ; Lise Getoor ; Dan Suciu
HIGHLIGHT:      In this paper, we present a formal framework for causal inference from such relational data.

29, TITLE:      Normalizing Flows with Multi-Scale Autoregressive Priors
http://arxiv.org/abs/2004.03891
AUTHORS:        Shweta Mahajan ; Apratim Bhattacharyya ; Mario Fritz ; Bernt Schiele ; Stefan Roth
COMMENTS:       To appear in CVPR 2020
HIGHLIGHT:      In this work, we improve the representational power of flow-based models by introducing channel-wise dependencies in their latent space through multi-scale autoregressive priors (mAR).

30, TITLE:      Constrained Multi-shape Evolution for Overlapping Cytoplasm Segmentation
http://arxiv.org/abs/2004.03892
AUTHORS:        Youyi Song ; Lei Zhu ; Baiying Lei ; Bin Sheng ; Qi Dou ; Jing Qin ; Kup-Sze Choi
COMMENTS:       12 pages and 6 figures
HIGHLIGHT:      In this paper, we present a novel and effective shape prior-based approach, called constrained multi-shape evolution, that segments all overlapping cytoplasms in the clump simultaneously by jointly evolving each cytoplasm's shape guided by the modeled shape priors.

31, TITLE:      Weakly Supervised Semantic Point Cloud Segmentation:Towards 10X Fewer Labels
http://arxiv.org/abs/2004.04091
AUTHORS:        Xun Xu ; Gim Hee Lee
COMMENTS:       CVPR2020
HIGHLIGHT:      In this work, we propose a weakly supervised point cloud segmentation approach which requires only a tiny fraction of points to be labelled in the training stage.

32, TITLE:      Beyond Photometric Consistency: Gradient-based Dissimilarity for Improving Visual Odometry and Stereo Matching
http://arxiv.org/abs/2004.04090
AUTHORS:        Jan Quenzel ; Radu Alexandru Rosu ; Thomas Läbe ; Cyrill Stachniss ; Sven Behnke
COMMENTS:       Accepted for International Conference on Robotic and Automation (ICRA), 2020
HIGHLIGHT:      In this paper, we investigate a new metric for registering images that builds upon on the idea of the photometric error.

33, TITLE:      Time accelerated image super-resolution using shallow residual feature representative network
http://arxiv.org/abs/2004.04093
AUTHORS:        Meenu Ajith ; Aswathy Rajendra Kurup ; Manel Martínez-Ramón
HIGHLIGHT:      To alleviate these issues, we developed an innovative shallow residual feature representative network (SRFRN) that uses a bicubic interpolated low-resolution image as input and residual representative units (RFR) which include serially stacked residual non-linear convolutions.

34, TITLE:      Satellite-based Prediction of Forage Conditions for Livestock in Northern Kenya
http://arxiv.org/abs/2004.04081
AUTHORS:        Andrew Hobbs ; Stacey Svetlichnaya
HIGHLIGHT:      This paper introduces the first dataset of satellite images labeled with forage quality by on-the-ground experts and provides proof of concept for applying computer vision methods to index-based drought insurance.

35, TITLE:      Explicit Reordering for Neural Machine Translation
http://arxiv.org/abs/2004.03818
AUTHORS:        Kehai Chen ; Rui Wang ; Masao Utiyama ; Eiichiro Sumita
HIGHLIGHT:      Thus, we propose a novel reordering method to explicitly model this reordering information for the Transformer-based NMT.

36, TITLE:      MirrorNet: A Deep Bayesian Approach to Reflective 2D Pose Estimation from Human Images
http://arxiv.org/abs/2004.03811
AUTHORS:        Takayuki Nakatsuka ; Kazuyoshi Yoshii ; Yuki Koyama ; Satoru Fukayama ; Masataka Goto ; Shigeo Morishima
COMMENTS:       19 pages
HIGHLIGHT:      This paper proposes a statistical approach to 2D pose estimation from human images.

37, TITLE:      Feature Re-Learning with Data Augmentation for Video Relevance Prediction
http://arxiv.org/abs/2004.03815
AUTHORS:        Jianfeng Dong ; Xun Wang ; Leimin Zhang ; Chaoxi Xu ; Gang Yang ; Xirong Li
COMMENTS:       accepted by IEEE Transactions on Knowledge and Data Engineering (TKDE)
HIGHLIGHT:      We propose in this paper feature re-learning for improving video relevance prediction, with no need of revisiting the original video content.

38, TITLE:      SciWING -- A Software Toolkit for Scientific Document Processing
http://arxiv.org/abs/2004.03807
AUTHORS:        Abhinav Ramesh Kashyap ; Min-Yen Kan
COMMENTS:       6 pages, 3 figures
HIGHLIGHT:      We introduce SciWING, an open-source software toolkit which provides access to pre-trained models for scientific document processing tasks, inclusive of citation string parsing and logical structure recovery.

39, TITLE:      Improving BERT with Self-Supervised Attention
http://arxiv.org/abs/2004.03808
AUTHORS:        Xiaoyu Kou ; Yaming Yang ; Yujing Wang ; Ce Zhang ; Yiren Chen ; Yunhai Tong ; Yan Zhang ; Jing Bai
HIGHLIGHT:      In this paper, we propose a novel technique, called Self-Supervised Attention (SSA) to help facilitate this generalization challenge.

40, TITLE:      Multi-Agent Task-Oriented Dialog Policy Learning with Role-Aware Reward Decomposition
http://arxiv.org/abs/2004.03809
AUTHORS:        Ryuichi Takanobu ; Runze Liang ; Minlie Huang
COMMENTS:       ACL 2020 long paper
HIGHLIGHT:      To avoid explicitly building a user simulator beforehand, we propose Multi-Agent Dialog Policy Learning, which regards both the system and the user as the dialog agents.

41, TITLE:      HybridDNN: A Framework for High-Performance Hybrid DNN Accelerator Design and Implementation
http://arxiv.org/abs/2004.03804
AUTHORS:        Hanchen Ye ; Xiaofan Zhang ; Zhize Huang ; Gengsheng Chen ; Deming Chen
COMMENTS:       Published as a conference paper at Design Automation Conference 2020 (DAC'20)
HIGHLIGHT:      To speedup Deep Neural Networks (DNN) accelerator design and enable effective implementation, we propose HybridDNN, a framework for building high-performance hybrid DNN accelerators and delivering FPGA-based hardware implementations.

42, TITLE:      State of the Art on Neural Rendering
http://arxiv.org/abs/2004.03805
AUTHORS:        Ayush Tewari ; Ohad Fried ; Justus Thies ; Vincent Sitzmann ; Stephen Lombardi ; Kalyan Sunkavalli ; Ricardo Martin-Brualla ; Tomas Simon ; Jason Saragih ; Matthias Nießner ; Rohit Pandey ; Sean Fanello ; Gordon Wetzstein ; Jun-Yan Zhu ; Christian Theobalt ; Maneesh Agrawala ; Eli Shechtman ; Dan B Goldman ; Michael Zollhöfer
COMMENTS:       Eurographics 2020 survey paper
HIGHLIGHT:      We focus on approaches that combine classic computer graphics techniques with deep generative models to obtain controllable and photo-realistic outputs.

43, TITLE:      Putting a Spin on Language: A Quantum Interpretation of Unary Connectives for Linguistic Applications
http://arxiv.org/abs/2004.04128
AUTHORS:        A. D. Correia ; H. T. C. Stoof ; M. Moortgat
HIGHLIGHT:      Our method introduces a way of simultaneously representing co-existing interpretations of ambiguous utterances, and provides a uniform framework for the integration of lexical and derivational ambiguity.

44, TITLE:      A computational theoretical approach for mining data on transient events from databases of high energy astrophysics experiments
http://arxiv.org/abs/2004.04131
AUTHORS:        Francesco Lazzarotto ; Marco Feroci ; Maria Teresa Pazienza
COMMENTS:       9 pages, 6 figures (in colors). Conference poster at Third Rome Workshop on Gamma-Ray Bursts in the Afterglow Era, Held 27-30 September 2002 at CNR Headquarters, Rome, Italy
HIGHLIGHT:      We present a computational formal model to apply techniques of modern computer science -such as Data Mining (DM) and Knowledge Discovering in Databases (KDD)- to a generic, large database derived from a high energy astrophysics experiment.

45, TITLE:      CURL: Contrastive Unsupervised Representations for Reinforcement Learning
http://arxiv.org/abs/2004.04136
AUTHORS:        Aravind Srinivas ; Michael Laskin ; Pieter Abbeel
COMMENTS:       First two authors contributed equally, website: https://mishalaskin.github.io/curl code: https://github.com/MishaLaskin/curl
HIGHLIGHT:      We present CURL: Contrastive Unsupervised Representations for Reinforcement Learning.

46, TITLE:      Entity-Switched Datasets: An Approach to Auditing the In-Domain Robustness of Named Entity Recognition Models
http://arxiv.org/abs/2004.04123
AUTHORS:        Oshin Agarwal ; Yinfei Yang ; Byron C. Wallace ; Ani Nenkova
HIGHLIGHT:      We propose a method for auditing the in-domain robustness of systems, focusing specifically on differences in performance due to the national origin of entities. We create entity-switched datasets, in which named entities in the original texts are replaced by plausible named entities of the same type but of different national origin.

47, TITLE:      Skin Diseases Detection using LBP and WLD- An Ensembling Approach
http://arxiv.org/abs/2004.04122
AUTHORS:        Arnab Banerjee ; Nibaran Das ; Mita Nasipuri
HIGHLIGHT:      Considering the importance, we propose an automatic technique to detect three popular skin diseases- Leprosy, Tinea versicolor and Vitiligofrom the images of skin lesions.

48, TITLE:      LadaBERT: Lightweight Adaptation of BERT through Hybrid Model Compression
http://arxiv.org/abs/2004.04124
AUTHORS:        Yihuan Mao ; Yujing Wang ; Chufan Wu ; Chen Zhang ; Yang Wang ; Yaming Yang ; Quanlu Zhang ; Yunhai Tong ; Jing Bai
HIGHLIGHT:      In this paper, we address this issue by proposing a hybrid solution named LadaBERT (Lightweight adaptation of BERT through hybrid model compression), which combines the advantages of different model compression methods, including weight pruning, matrix factorization and knowledge distillation.

49, TITLE:      Self-Supervised Monocular Scene Flow Estimation
http://arxiv.org/abs/2004.04143
AUTHORS:        Junhwa Hur ; Stefan Roth
COMMENTS:       To appear at CVPR 2020 (Oral)
HIGHLIGHT:      We propose a novel monocular scene flow method that yields competitive accuracy and real-time performance.

50, TITLE:      Slicing and dicing soccer: automatic detection ofcomplex events from spatio-temporal data
http://arxiv.org/abs/2004.04147
AUTHORS:        Lia Morra ; Francesco Manigrasso ; Giuseppe Canto ; Claudio Gianfrate ; Enrico Guarino ; Fabrizio Lamberti
COMMENTS:       accepted at 17th International Conference on Image Analysis and Recognition ICIAR 2020
HIGHLIGHT:      This paper presents a comprehensive approach for de-tecting a wide range of complex events in soccer videos starting frompositional data.

51, TITLE:      Empirical Perspectives on One-Shot Semi-supervised Learning
http://arxiv.org/abs/2004.04141
AUTHORS:        Leslie N. Smith ; Adam Conovaloff
COMMENTS:       Short paper with interesting results pointing to further investigation
HIGHLIGHT:      We empirically investigate the scenario where one has access to large amounts of unlabeled data but require labeling only a single prototypical sample per class in order to train a deep network (i.e., one-shot semi-supervised learning).

52, TITLE:      Semantic Image Manipulation Using Scene Graphs
http://arxiv.org/abs/2004.03677
AUTHORS:        Helisa Dhamo ; Azade Farshad ; Iro Laina ; Nassir Navab ; Gregory D. Hager ; Federico Tombari ; Christian Rupprecht
COMMENTS:       CVPR 2020
HIGHLIGHT:      Our goal is to encode image information in a given constellation and from there on generate new constellations, such as replacing objects or even changing relationships between objects, while respecting the semantics and style from the original image.

53, TITLE:      Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?
http://arxiv.org/abs/2004.03685
AUTHORS:        Alon Jacovi ; Yoav Goldberg
COMMENTS:       Accepted to ACL 2020
HIGHLIGHT:      We call for more clearly differentiating between different desired criteria an interpretation should satisfy, and focus on the faithfulness criteria.

54, TITLE:      Exemplar Fine-Tuning for 3D Human Pose Fitting Towards In-the-Wild 3D Human Pose Estimation
http://arxiv.org/abs/2004.03686
AUTHORS:        Hanbyul Joo ; Natalia Neverova ; Andrea Vedaldi
HIGHLIGHT:      We propose a method for building large collections of human poses with full 3D annotations captured `in the wild', for which specialized capture equipment cannot be used.

55, TITLE:      Radon cumulative distribution transform subspace modeling for image classification
http://arxiv.org/abs/2004.03669
AUTHORS:        Mohammad Shifat-E-Rabbi ; Xuwang Yin ; Abu Hasnat Mohammad Rubaiyat ; Shiying Li ; Soheil Kolouri ; Akram Aldroubi ; Jonathan M. Nichols ; Gustavo K. Rohde
COMMENTS:       14 pages, 11 figures
HIGHLIGHT:      We present a new supervised image classification method for problems where the data at hand conform to certain deformation models applied to unknown prototypes or templates.

56, TITLE:      Dynamic Data Selection and Weighting for Iterative Back-Translation
http://arxiv.org/abs/2004.03672
AUTHORS:        Zi-Yi Dou ; Antonios Anastasopoulos ; Graham Neubig
HIGHLIGHT:      In this paper, we provide insights into this commonly used approach and generalize it to a dynamic curriculum learning strategy, which is applied to iterative back-translation models.

57, TITLE:      Spatio-temporal Learning from Longitudinal Data for Multiple Sclerosis Lesion Segmentation
http://arxiv.org/abs/2004.03675
AUTHORS:        Stefan Denner ; Ashkan Khakzar ; Moiz Sajid ; Mahdi Saleh ; Ziga Spiclin ; Seong Tae Kim ; Nassir Navab
HIGHLIGHT:      To that end, we propose two approaches: Our longitudinal segmentation architecture which is grounded upon early-fusion of longitudinal data.

58, TITLE:      SA-UNet: Spatial Attention U-Net for Retinal Vessel Segmentation
http://arxiv.org/abs/2004.03696
AUTHORS:        Changlu Guo ; Márton Szemenyei ; Yugen Yi ; Wenle Wang ; Buer Chen ; Changqi Fan
COMMENTS:       Submitted to IEEE ICPR 2020
HIGHLIGHT:      In this work, we propose a lightweight network named Spatial Attention U-Net (SA-UNet) that does not require thousands of annotated training samples and can be utilized in a data augmentation manner to use the available annotated samples more efficiently.

59, TITLE:      Dense Residual Network for Retinal Vessel Segmentation
http://arxiv.org/abs/2004.03697
AUTHORS:        Changlu Guo ; Márton Szemenyei ; Yugen Yi ; Ying Xue ; Wei Zhou ; Yangyuan Li
COMMENTS:       Accepted by IEEE ICASSP 2020
HIGHLIGHT:      In this work, we propose an efficient method to segment blood vessels in Scanning Laser Ophthalmoscopy (SLO) retinal images.

60, TITLE:      Coronavirus (COVID-19) Classification using Deep Features Fusion and Ranking Technique
http://arxiv.org/abs/2004.03698
AUTHORS:        Umut Ozkaya ; Saban Ozturk ; Mucahid Barstugan
COMMENTS:       13 pages, 6 figures
HIGHLIGHT:      In this study, a novel method was proposed as fusing and ranking deep features to detect COVID-19 in early phase.

61, TITLE:      Frequency, Acceptability, and Selection: A case study of clause-embedding
http://arxiv.org/abs/2004.04106
AUTHORS:        Aaron Steven White ; Kyle Rawlins
HIGHLIGHT:      We investigate the relationship between the frequency with which verbs are found in particular subcategorization frames and the acceptability of those verbs in those frames, focusing in particular on subordinate clause-taking verbs, such as "think", "want", and "tell".

62, TITLE:      Reservoir Computing using High Order Synchronization of Coupled Oscillators
http://arxiv.org/abs/2004.04114
AUTHORS:        A. A. Velichko ; D. V. Ryabokon ; S. D. Khanin ; A. V. Sidorenko ; A. G. Rikkiev
COMMENTS:       8 pages, 7 figures
HIGHLIGHT:      We propose a concept for reservoir computing on oscillators using the high-order synchronization effect.

63, TITLE:      KdConv: A Chinese Multi-domain Dialogue Dataset Towards Multi-turn Knowledge-driven Conversation
http://arxiv.org/abs/2004.04100
AUTHORS:        Hao Zhou ; Chujie Zheng ; Kaili Huang ; Minlie Huang ; Xiaoyan Zhu
HIGHLIGHT:      In this paper, we propose a Chinese multi-domain knowledge-driven conversation dataset, KdConv, which grounds the topics in multi-turn conversations to knowledge graphs.

64, TITLE:      Cross-lingual Emotion Intensity Prediction
http://arxiv.org/abs/2004.04103
AUTHORS:        Irean Navas Alejo ; Toni Badia ; Jeremy Barnes
HIGHLIGHT:      Consequently, we explore cross-lingual transfer approaches for fine-grained emotion detection in Spanish and Catalan tweets. To this end we annotate a test set of Spanish and Catalan tweets using Best-Worst scaling.

65, TITLE:      Mobile-Based Deep Learning Models for Banana Diseases Detection
http://arxiv.org/abs/2004.03718
AUTHORS:        Sophia Sanga ; Victor Mero ; Dina Machuve ; Davis Mwanganda
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      We used a dataset of 3000 banana leaves images.

66, TITLE:      File Classification Based on Spiking Neural Networks
http://arxiv.org/abs/2004.03953
AUTHORS:        Ana Stanojevic ; Giovanni Cherubini ; Timoleon Moraitis ; Abu Sebastian
COMMENTS:       5 pages. 5 figures. Accepted at ISCAS 2020 for publication
HIGHLIGHT:      In this paper, we propose a system for file classification in large data sets based on spiking neural networks (SNNs).

67, TITLE:      A Polynomial Neural Network with Controllable Precision and Human-Readable Topology for Prediction and System Identification
http://arxiv.org/abs/2004.03955
AUTHORS:        Gang Liu ; Jing Wang
COMMENTS:       Submitted it to IEEE TNNLS; 8 pages, 9 figures
HIGHLIGHT:      This paper presents a controllable and readable polynomial neural network (CR-PNN) for approximation, prediction, and system identification.

68, TITLE:      Compiling Spiking Neural Networks to Neuromorphic Hardware
http://arxiv.org/abs/2004.03717
AUTHORS:        Shihao Song ; Adarsha Balaji ; Anup Das ; Nagarajan Kandasamy ; James Shackleford
COMMENTS:       10 pages, 17 figures, accepted at 21st ACM SIGPLAN/SIGBED International Conference on Languages, Compilers, and Tools for Embedded Systems (LCTES 2020)
HIGHLIGHT:      We propose an approach to analyze and compile SNNs on a resource-constrained neuromorphic hardware, providing guarantee on key performance metrics such as execution time and throughput.

69, TITLE:      Adversary Helps: Gradient-based Device-Free Domain-Independent Gesture Recognition
http://arxiv.org/abs/2004.03961
AUTHORS:        Jianwei Liu ; Jinsong Han ; Feng Lin ; Kui Ren
HIGHLIGHT:      Though some solutions, such as adversarial learning, transfer learning and body-coordinate velocity profile, have been proposed to achieve cross-domain recognition, these solutions more or less have flaws.

70, TITLE:      Byte Pair Encoding is Suboptimal for Language Model Pretraining
http://arxiv.org/abs/2004.03720
AUTHORS:        Kaj Bostrom ; Greg Durrett
COMMENTS:       4 pages, 2 figures
HIGHLIGHT:      Across downstream tasks, we find that the unigram LM tokenization method consistently matches or outperforms BPE.

71, TITLE:      Context-Aware Group Captioning via Self-Attention and Contrastive Features
http://arxiv.org/abs/2004.03708
AUTHORS:        Zhuowan Li ; Quan Tran ; Long Mai ; Zhe Lin ; Alan Yuille
COMMENTS:       To appear in CVPR 2020; Project page: https://lizw14.github.io/project/groupcap
HIGHLIGHT:      To solve this problem, we propose a framework combining self-attention mechanism with contrastive feature construction to effectively summarize common information from each image group while capturing discriminative information between them.

72, TITLE:      Channel Attention Residual U-Net for Retinal Vessel Segmentation
http://arxiv.org/abs/2004.03702
AUTHORS:        Changlu Guo ; Márton Szemenyei ; Yugen Yi ; Wei Zhou
COMMENTS:       Submitted to IEEE ICIP 2020
HIGHLIGHT:      In this work, we propose a new deep learning model, namely Channel Attention Residual U-Net (CAR-U-Net), to accurately segment retinal vascular and non-vascular pixels.

73, TITLE:      Improved YOLOv3 Object Classification in Intelligent Transportation System
http://arxiv.org/abs/2004.03948
AUTHORS:        Yang Zhang ; Changhui Hu ; Xiaobo Lu
HIGHLIGHT:      In this paper, an algorithm based on YOLOv3 is proposed to realize the detection and classification of vehicles, drivers, and people on the highway, so as to achieve the purpose of distinguishing driver and passenger and form a one-to-one correspondence between vehicles and drivers.

74, TITLE:      Long-Tailed Recognition Using Class-Balanced Experts
http://arxiv.org/abs/2004.03706
AUTHORS:        Saurabh Sharma ; Ning Yu ; Mario Fritz ; Bernt Schiele
COMMENTS:       18 pages, 5 figures, 9 tables
HIGHLIGHT:      In this work we address the problem of long tail recognition wherein the training set is highly imbalanced and the test set is kept balanced.

75, TITLE:      Pre-training is a Hot Topic: Contextualized Document Embeddings Improve Topic Coherence
http://arxiv.org/abs/2004.03974
AUTHORS:        Federico Bianchi ; Silvia Terragni ; Dirk Hovy
HIGHLIGHT:      We combine pre-trained representations and neural topic models.

76, TITLE:      Locality Preserving Loss to Align Vector Spaces
http://arxiv.org/abs/2004.03734
AUTHORS:        Ashwinkumar Ganesan ; Frank Ferraro ; Tim Oates
HIGHLIGHT:      We present a locality preserving loss (LPL)that improves the alignment between vector space representations (i.e., word or sentence embeddings) while separating (increasing distance between) uncorrelated representations as compared to the standard method that minimizes the mean squared error (MSE) only.

77, TITLE:      Learning to Detect Head Movement in Unconstrained Remote Gaze Estimation in the Wild
http://arxiv.org/abs/2004.03737
AUTHORS:        Zhecan Wang ; Jian Zhao ; Cheng Lu ; Han Huang ; Fan Yang ; Lianji Li ; Yandong Guo
COMMENTS:       2020 Winter Conference on Applications of Computer Vision
HIGHLIGHT:      In this paper, we propose novel end-to-end appearance-based gaze estimation methods that could more robustly incorporate different levels of head-pose representations into gaze estimation. To better demonstrate the advantage of our methods, we further propose a new benchmark dataset with the most rich distribution of head-gaze combination reflecting real-world scenarios.

78, TITLE:      Towards Evaluating the Robustness of Chinese BERT Classifiers
http://arxiv.org/abs/2004.03742
AUTHORS:        Boxin Wang ; Boyuan Pan ; Xin Li ; Bo Li
HIGHLIGHT:      In this paper, we show that, however, such BERT-based models are vulnerable under character-level adversarial attacks.

79, TITLE:      Conditional Rap Lyrics Generation with Denoising Autoencoders
http://arxiv.org/abs/2004.03965
AUTHORS:        Nikola I. Nikolov ; Eric Malmi ; Curtis G. Northcutt ; Loreto Parisi
HIGHLIGHT:      We develop a method for automatically synthesizing a rap verse given an input text written in another form, such as a summary of a news article.

80, TITLE:      Learning 3D Semantic Scene Graphs from 3D Indoor Reconstructions
http://arxiv.org/abs/2004.03967
AUTHORS:        Johanna Wald ; Helisa Dhamo ; Nassir Navab ; Federico Tombari
COMMENTS:       first two authors contributed equally, CVPR 2020, video https://youtu.be/8D3HjYf6cYw
HIGHLIGHT:      In particular, we propose a learned method that regresses a scene graph from the point cloud of a scene.

81, TITLE:      Understanding Knowledge Gaps in Visual Question Answering: Implications for Gap Identification and Testing
http://arxiv.org/abs/2004.03755
AUTHORS:        Goonmeet Bajaj ; Bortik Bandyopadhyay ; Daniel Schmidt ; Pranav Maneriker ; Christopher Myers ; Srinivasan Parthasarathy
HIGHLIGHT:      In order to reduce the skew in the distribution of questions across KGs, we introduce a targeted question generation model.

82, TITLE:      DashCam Pay: A System for In-vehicle Payments Using Face and Voice
http://arxiv.org/abs/2004.03756
AUTHORS:        Cori Tymoszek ; Sunpreet S. Arora ; Kim Wagner ; Anil K. Jain
COMMENTS:       9 pages, 6 figures
HIGHLIGHT:      We present an open loop system, called DashCam Pay, that enables in-vehicle payments using face and voice biometrics. For initial feasibility analysis, we collected data from 20 different subjects at two different sites using a commercially available dashcam, and evaluated open-source biometric algorithms on the collected data.

83, TITLE:      DialBERT: A Hierarchical Pre-Trained Model for Conversation Disentanglement
http://arxiv.org/abs/2004.03760
AUTHORS:        Tianda Li ; Jia-Chen Gu ; Xiaodan Zhu ; Quan Liu ; Zhen-Hua Ling ; Zhiming Su ; Si Wei
HIGHLIGHT:      We propose a new model, named Dialogue BERT (DialBERT), which integrates local and global semantics in a single stream of messages to disentangle the conversations that mixed together.

84, TITLE:      Adaptive Transformers in RL
http://arxiv.org/abs/2004.03761
AUTHORS:        Shakti Kumar ; Jerrod Parker ; Panteha Naderian
COMMENTS:       10 pages with 9 figures and 4 tables. Main text is 6 pages, appendix is 4 pages
HIGHLIGHT:      In this work we first partially replicate the results shown in Stabilizing Transformers in RL on both reactive and memory based environments.

85, TITLE:      Generating Narrative Text in a Switching Dynamical System
http://arxiv.org/abs/2004.03762
AUTHORS:        Noah Weber ; Leena Shekhar ; Heeyoung Kwon ; Niranjan Balasubramanian ; Nathanael Chambers
HIGHLIGHT:      This paper introduces a new model that integrates explicit narrative structure with neural language models, formalizing narrative modeling as a Switching Linear Dynamical System (SLDS).

86, TITLE:      e-SNLI-VE-2.0: Corrected Visual-Textual Entailment with Natural Language Explanations
http://arxiv.org/abs/2004.03744
AUTHORS:        Virginie Do ; Oana-Maria Camburu ; Zeynep Akata ; Thomas Lukasiewicz
HIGHLIGHT:      e-SNLI-VE-2.0: Corrected Visual-Textual Entailment with Natural Language Explanations

87, TITLE:      Multi-Person Absolute 3D Human Pose Estimation with Weak Depth Supervision
http://arxiv.org/abs/2004.03989
AUTHORS:        Marton Veges ; Andras Lorincz
HIGHLIGHT:      To mitigate this issue, we introduce a network that can be trained with additional RGB-D images in a weakly supervised fashion.

88, TITLE:      COVID_MTNet: COVID-19 Detection with Multi-Task Deep Learning Approaches
http://arxiv.org/abs/2004.03747
AUTHORS:        Md Zahangir Alom ; M M Shaifur Rahman ; Mst Shamima Nasrin ; Tarek M. Taha ; Vijayan K. Asari
COMMENTS:       11 pages, 15 figures
HIGHLIGHT:      In this paper, we propose a fast and efficient way to identify COVID-19 patients with multi-task deep learning (DL) methods.

89, TITLE:      Downstream Model Design of Pre-trained Language Model for Relation Extraction Task
http://arxiv.org/abs/2004.03786
AUTHORS:        Cheng Li ; Ye Tian
HIGHLIGHT:      In this paper, a new network architecture with a special loss function is designed to serve as a downstream model of PLMs for supervised relation extraction.

90, TITLE:      Deep Adaptive Inference Networks for Single Image Super-Resolution
http://arxiv.org/abs/2004.03915
AUTHORS:        Ming Liu ; Zhilu Zhang ; Liya Hou ; Wangmeng Zuo ; Lei Zhang
COMMENTS:       Code can be found at https://github.com/csmliu/AdaDSR
HIGHLIGHT:      In this paper, we take a step forward to address this issue by leveraging the adaptive inference networks for deep SISR (AdaDSR).

91, TITLE:      Which one is the dax? Achieving mutual exclusivity with neural networks
http://arxiv.org/abs/2004.03902
AUTHORS:        Kristina Gulordava ; Thomas Brochhagen ; Gemma Boleda
HIGHLIGHT:      In this study, we investigate whether and under which circumstances neural models can exhibit analogous behavior.

92, TITLE:      Image super-resolution reconstruction based on attention mechanism and feature fusion
http://arxiv.org/abs/2004.03939
AUTHORS:        Jiawen Lyn ; Sen Yan
HIGHLIGHT:      Experimental results show that the proposed method can achieve better performance over other representative super-resolution reconstruction algorithms in objective quantitative metrics and visual quality.

93, TITLE:      Densities of almost-surely terminating probabilistic programs are differentiable almost everywhere
http://arxiv.org/abs/2004.03924
AUTHORS:        Carol Mak ; C. -H. Luke Ong ; Hugo Paquet ; Dominik Wagner
HIGHLIGHT:      We study the differential properties of higher-order statistical probabilistic programs with recursion and conditioning.

94, TITLE:      Word frequency and sentiment analysis of twitter messages during Coronavirus pandemic
http://arxiv.org/abs/2004.03925
AUTHORS:        Nikhil Kumar Rajput ; Bhavya Ahuja Grover ; Vipin Kumar Rathi
HIGHLIGHT:      This paper presents a statistical analysis of the twitter messages related to this disease posted since January 2020.

95, TITLE:      Exploiting Redundancy in Pre-trained Language Models for Efficient Transfer Learning
http://arxiv.org/abs/2004.04010
AUTHORS:        Fahim Dalvi ; Hassan Sajjad ; Nadir Durrani ; Yonatan Belinkov
COMMENTS:       11 Pages, 5 figures
HIGHLIGHT:      We propose a novel feature selection method, which takes advantage of these redundancies to reduce the size of the pre-trained features.

96, TITLE:      Learning from Learners: Adapting Reinforcement Learning Agents to be Competitive in a Card Game
http://arxiv.org/abs/2004.04000
AUTHORS:        Pablo Barros ; Ana Tanevska ; Alessandra Sciutti
COMMENTS:       Submitted to ICPR2020
HIGHLIGHT:      In this paper, we present a broad study on how popular reinforcement learning algorithms can be adapted and implemented to learn and to play a real-world implementation of a competitive multiplayer card game.

97, TITLE:      Transfer learning and subword sampling for asymmetric-resource one-to-many neural translation
http://arxiv.org/abs/2004.04002
AUTHORS:        Stig-Arne Grönroos ; Sami Virpioja ; Mikko Kurimo
COMMENTS:       26 pages, 12 tables, 7 figures. Submitted (Mar 2020) to the Machine Translation journal Special Issue on Machine Translation for Low-Resource Languages (Springer)
HIGHLIGHT:      We review these approaches in the context of an asymmetric-resource one-to-many translation task, in which the pair of target languages are related, with one being a very low-resource and the other a higher-resource language.

98, TITLE:      New Opportunities for the Formal Proof of Computational Real Geometry?
http://arxiv.org/abs/2004.04034
AUTHORS:        Erika {Á}brahám ; James Davenport ; Matthew England ; Gereon Kremer ; Zak Tonks
HIGHLIGHT:      The purpose of this paper is to explore the question "to what extent could we produce formal, machine-verifiable, proofs in real algebraic geometry?"

99, TITLE:      DynaBERT: Dynamic BERT with Adaptive Width and Depth
http://arxiv.org/abs/2004.04037
AUTHORS:        Lu Hou ; Lifeng Shang ; Xin Jiang ; Qun Liu
HIGHLIGHT:      In this paper, we propose a novel dynamic BERT model (abbreviated as DynaBERT), which can run at adaptive width and depth.

100, TITLE:     Neural Networks Model for Travel Time Prediction Based on ODTravel Time Matrix
http://arxiv.org/abs/2004.04030
AUTHORS:        Ayobami E. Adewale ; Amnir Hadachi
HIGHLIGHT:      In this study, two neural network models namely multi-layer(MLP) perceptron and long short-term model(LSTM) are developed for predicting link travel time of a busy route with input generated using Origin-Destination travel time matrix derived from a historical GPS dataset.

101, TITLE:     A Deep Learning Approach for Determining Effects of Tuta Absoluta in Tomato Plants
http://arxiv.org/abs/2004.04023
AUTHORS:        Denis P. Rubanga ; Loyani K. Loyani ; Mgaya Richard ; Sawahiko Shimada
COMMENTS:       Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)
HIGHLIGHT:      In this study, we propose a Convolutional Neural Network (CNN) approach in determining the effects of Tuta absoluta in tomato plants.

102, TITLE:     Continual Learning with Gated Incremental Memories for sequential data processing
http://arxiv.org/abs/2004.04077
AUTHORS:        Andrea Cossu ; Antonio Carta ; Davide Bacciu
COMMENTS:       Accepted as a conference paper at 2020 International Joint Conference on Neural Networks (IJCNN 2020). Part of 2020 IEEE World Congress on Computational Intelligence (IEEE WCCI 2020)
HIGHLIGHT:      This work proposes a Recurrent Neural Network (RNN) model for CL that is able to deal with concept drift in input distribution without forgetting previously acquired knowledge.

103, TITLE:     Are All Good Word Vector Spaces Isomorphic?
http://arxiv.org/abs/2004.04070
AUTHORS:        Ivan Vulić ; Sebastian Ruder ; Anders Søgaard
COMMENTS:       Equal contribution from all three authors
HIGHLIGHT:      In this work, we ask whether non-isomorphism is also crucially a sign of degenerate word vector spaces.

104, TITLE:     Convolutional neural net face recognition works in non-human-like ways
http://arxiv.org/abs/2004.04069
AUTHORS:        P. J. B. Hancock ; R. S. Somai ; V. R. Mileva ; A. K. Bobak
COMMENTS:       8 pages, 2 figures. Submitted to Royal Society Open Science
HIGHLIGHT:      We tested six commercial face recognition CNNs and found that they outperform typical human participants on standard face matching tasks.

105, TITLE:     Self-Attention Gazetteer Embeddings for Named-Entity Recognition
http://arxiv.org/abs/2004.04060
AUTHORS:        Stanislav Peshterliev ; Christophe Dupuy ; Imre Kiss
COMMENTS:       Preprint
HIGHLIGHT:      In this work, we present GazSelfAttn, a novel gazetteer embedding approach that uses self-attention and match span encoding to build enhanced gazetteer embeddings.

106, TITLE:     Satirical News Detection with Semantic Feature Extraction and Game-theoretic Rough Sets
http://arxiv.org/abs/2004.03788
AUTHORS:        Yue Zhou ; Yan Zhang ; JingTao Yao
COMMENTS:       12 pages
HIGHLIGHT:      To address these issues, we collect satirical and legitimate news tweets, and propose a semantic feature based approach.

107, TITLE:     Learning for Scale-Arbitrary Super-Resolution from Scale-Specific Networks
http://arxiv.org/abs/2004.03791
AUTHORS:        Longguang Wang ; Yingqian Wang ; Zaiping Lin ; Jungang Yang ; Wei An ; Yulan Guo
HIGHLIGHT:      In this paper, we propose to learn a scale-arbitrary image SR network from scale-specific networks.

108, TITLE:     CALM: Continuous Adaptive Learning for Language Modeling
http://arxiv.org/abs/2004.03794
AUTHORS:        Kristjan Arumae ; Parminder Bhatia
HIGHLIGHT:      In this work we propose CALM, Continuous Adaptive Learning for Language Modeling: techniques to render models which retain knowledge across multiple domains.

109, TITLE:     DMLO: Deep Matching LiDAR Odometry
http://arxiv.org/abs/2004.03796
AUTHORS:        Naiyan Wang ; Zhichao Li
HIGHLIGHT:      In this paper, we present Deep Matching LiDAR Odometry (DMLO), a novel learning-based framework which makes the feature matching method applicable to LiDAR odometry task.



==========Updates to Previous Papers==========
1, TITLE:       MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Mask
http://arxiv.org/abs/2003.10955
AUTHORS:        Shengyu Zhao ; Yilun Sheng ; Yue Dong ; Eric I-Chao Chang ; Yan Xu
COMMENTS:       CVPR 2020 (Oral)
HIGHLIGHT:      In this paper, we propose an asymmetric occlusion-aware feature matching module, which can learn a rough occlusion mask that filters useless (occluded) areas immediately after feature warping without any explicit supervision.

2, TITLE:       Potential Field: Interpretable and Unified Representation for Trajectory Prediction
http://arxiv.org/abs/1911.07414
AUTHORS:        Shan Su ; Cheng Peng ; Jianbo Shi ; Chiho Choi
HIGHLIGHT:      In this work, we borrow the notion of potential field from physics as an interpretable and unified representation to model all stimuli.

3, TITLE:       AeroRIT: A New Scene for Hyperspectral Image Analysis
http://arxiv.org/abs/1912.08178
AUTHORS:        Aneesh Rangnekar ; Nilay Mokashi ; Emmett Ientilucci ; Christopher Kanan ; Matthew J. Hoffman
COMMENTS:       To appear in IEEE TGRS
HIGHLIGHT:      We investigate applying convolutional neural network (CNN) architecture to facilitate aerial hyperspectral scene understanding and present a new hyperspectral dataset-AeroRIT-that is large enough for CNN training.

4, TITLE:       XPersona: Evaluating Multilingual Personalized Chatbot
http://arxiv.org/abs/2003.07568
AUTHORS:        Zhaojiang Lin ; Zihan Liu ; Genta Indra Winata ; Samuel Cahyawijaya ; Andrea Madotto ; Yejin Bang ; Etsuko Ishii ; Pascale Fung
COMMENTS:       Preprint, 23 pages
HIGHLIGHT:      In this paper, we propose a multi-lingual extension of Persona-Chat, namely XPersona.

5, TITLE:       Encoding Visual Attributes in Capsules for Explainable Medical Diagnoses
http://arxiv.org/abs/1909.05926
AUTHORS:        Rodney LaLonde ; Drew Torigian ; Ulas Bagci
COMMENTS:       Paper is currently under review at an unnamed conference. Please respect double blind submissions. This paper will be updated with public code and any feedback from the peer-review process
HIGHLIGHT:      In this study, we introduce a novel multi-task learning algorithm based on capsule network to encode visual attributes towards image-based diagnosis.

6, TITLE:       Dynamically Pruned Message Passing Networks for Large-Scale Knowledge Graph Reasoning
http://arxiv.org/abs/1909.11334
AUTHORS:        Xiaoran Xu ; Wei Feng ; Yunsheng Jiang ; Xiaohui Xie ; Zhiqing Sun ; Zhi-Hong Deng
COMMENTS:       ICLR2020
HIGHLIGHT:      We propose Dynamically Pruned Message Passing Networks (DPMPN) for large-scale knowledge graph reasoning.

7, TITLE:       Gradient Centralization: A New Optimization Technique for Deep Neural Networks
http://arxiv.org/abs/2004.01461
AUTHORS:        Hongwei Yong ; Jianqiang Huang ; Xiansheng Hua ; Lei Zhang
COMMENTS:       20 pages, 7 figures, conference
HIGHLIGHT:      Different from these existing methods that mostly operate on activations or weights, we present a new optimization technique, namely gradient centralization (GC), which operates directly on gradients by centralizing the gradient vectors to have zero mean.

8, TITLE:       A Deep Learning Framework for Detection of Targets in Thermal Images to Improve Firefighting
http://arxiv.org/abs/1910.03617
AUTHORS:        Manish Bhattarai ; Manel Martínez-Ramón
HIGHLIGHT:      The objective of this research is to create an automated system that is capable of real-time, intelligent object detection and recognition and facilitates the improved situational awareness of firefighters during an emergency response.

9, TITLE:       Interactive Refinement of Cross-Lingual Word Embeddings
http://arxiv.org/abs/1911.03070
AUTHORS:        Michelle Yuan ; Mozhi Zhang ; Benjamin Van Durme ; Leah Findlater ; Jordan Boyd-Graber
COMMENTS:       First two authors contribute equally
HIGHLIGHT:      We introduce CLIME, an interactive system that allows a user to quickly refine cross-lingual word embeddings for a given classification problem.

10, TITLE:      FlapAI Bird: Training an Agent to Play Flappy Bird Using Reinforcement Learning Techniques
http://arxiv.org/abs/2003.09579
AUTHORS:        Tai Vu ; Leon Tran
COMMENTS:       typos corrected, references added
HIGHLIGHT:      We seek to apply reinforcement learning algorithms to the game Flappy Bird.

11, TITLE:      U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation
http://arxiv.org/abs/1907.10830
AUTHORS:        Junho Kim ; Minjae Kim ; Hyeonwoo Kang ; Kwanghee Lee
COMMENTS:       Accepted to ICLR 2020
HIGHLIGHT:      We propose a novel method for unsupervised image-to-image translation, which incorporates a new attention module and a new learnable normalization function in an end-to-end manner.

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

13, TITLE:      cGANs with Conditional Convolution Layer
http://arxiv.org/abs/1906.00709
AUTHORS:        Min-Cheol Sagong ; Yong-Goo Shin ; Yoon-Jae Yeo ; Seung Park ; Sung-Jea Ko
COMMENTS:       Submitted to IEEE Trans. Neural Networks and Learning Systems (TNNLS)
HIGHLIGHT:      In this paper, we propose a novel convolution layer, called the conditional convolution layer, which directly generates different feature maps by employing the weights which are adjusted depending on the conditions.

14, TITLE:      Mining Implicit Entity Preference from User-Item Interaction Data for Knowledge Graph Completion via Adversarial Learning
http://arxiv.org/abs/2003.12718
AUTHORS:        Gaole He ; Junyi Li ; Wayne Xin Zhao ; Peiju Liu ; Ji-Rong Wen
COMMENTS:       11 pages, 4 figures, 6 tables. Accepted as WWW 2020 paper
HIGHLIGHT:      In this paper, we take a new perspective that aims to leverage rich user-item interaction data (user interaction data for short) for improving the KGC task.

15, 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.

16, TITLE:      Shoestring: Graph-Based Semi-Supervised Learning with Severely Limited Labeled Data
http://arxiv.org/abs/1910.12976
AUTHORS:        Wanyu Lin ; Zhaolin Gao ; Baochun Li
COMMENTS:       9 pages, 5 tables, 3 figures, accepted at CVPR2020, source code will be released soon
HIGHLIGHT:      In this work, we advance this effective learning paradigm towards a scenario where labeled data are severely limited.

17, TITLE:      The complexity of simulating local measurements on quantum systems
http://arxiv.org/abs/1606.05626
AUTHORS:        Sevag Gharibian ; Justin Yirka
COMMENTS:       38 pages, 0 figures. Fixed bug in proof of Lemma 4.3 by extending Lemma 4.1 and redefining gamma' (see footnote 13)
HIGHLIGHT:      In this paper, we continue the study of P^QMA[log], obtaining the following lower and upper bounds.

18, TITLE:      Re-Translation Strategies For Long Form, Simultaneous, Spoken Language Translation
http://arxiv.org/abs/1912.03393
AUTHORS:        Naveen Arivazhagan ; Colin Cherry ; Te I ; Wolfgang Macherey ; Pallavi Baljekar ; George Foster
COMMENTS:       ICASSP 2020
HIGHLIGHT:      We investigate the problem of simultaneous machine translation of long-form speech content.

19, TITLE:      KorNLI and KorSTS: New Benchmark Datasets for Korean Natural Language Understanding
http://arxiv.org/abs/2004.03289
AUTHORS:        Jiyeon Ham ; Yo Joong Choe ; Kyubyong Park ; Ilji Choi ; Hyungjoon Soh
COMMENTS:       Access datasets at https://github.com/kakaobrain/KorNLUDatasets
HIGHLIGHT:      Motivated by this, we construct and release new datasets for Korean NLI and STS, dubbed KorNLI and KorSTS, respectively.

20, TITLE:      Learning Instance Occlusion for Panoptic Segmentation
http://arxiv.org/abs/1906.05896
AUTHORS:        Justin Lazarow ; Kwonjoon Lee ; Kunyu Shi ; Zhuowen Tu
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      To resolve this issue, we propose a branch that is tasked with modeling how two instance masks should overlap one another as a binary relation.

21, TITLE:      Autoencoders for Unsupervised Anomaly Segmentation in Brain MR Images: A Comparative Study
http://arxiv.org/abs/2004.03271
AUTHORS:        Christoph Baur ; Stefan Denner ; Benedikt Wiestler ; Shadi Albarqouni ; Nassir Navab
HIGHLIGHT:      The main principle behind these works is to learn a model of normal anatomy by learning to compress and recover healthy data.

22, TITLE:      AdversarialNAS: Adversarial Neural Architecture Search for GANs
http://arxiv.org/abs/1912.02037
AUTHORS:        Chen Gao ; Yunpeng Chen ; Si Liu ; Zhenxiong Tan ; Shuicheng Yan
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      In this paper, we propose an AdversarialNAS method specially tailored for Generative Adversarial Networks (GANs) to search for a superior generative model on the task of unconditional image generation.

23, TITLE:      Neural Architecture Search for Compressed Sensing Magnetic Resonance Image Reconstruction
http://arxiv.org/abs/2002.09625
AUTHORS:        Jiangpeng Yan ; Shuo Chen ; Xiu Li ; Yongbing Zhang
COMMENTS:       10 pages, submitted to Elsewier
HIGHLIGHT:      In this manuscript, we proposed a novel and efficient MR image reconstruction framework by Neural Architecture Search (NAS) algorithm.

24, TITLE:      The Ladder Algorithm: Finding Repetitive Structures in Medical Images by Induction
http://arxiv.org/abs/2001.11284
AUTHORS:        Rhydian Windsor ; Amir Jamaludin
COMMENTS:       5 pages, 4 figures, IEEE International Symposium on Biomedical Imaging (ISBI) 2020. Presentation: https://www.youtube.com/watch?v=khlBqpNGRnE
HIGHLIGHT:      In this paper we introduce the Ladder Algorithm; a novel recurrent algorithm to detect repetitive structures in natural images with high accuracy using little training data.

25, TITLE:      Orthogonal Convolutional Neural Networks
http://arxiv.org/abs/1911.12207
AUTHORS:        Jiayun Wang ; Yubei Chen ; Rudrasis Chakraborty ; Stella X. Yu
COMMENTS:       To appear in CVPR 2020, project page: http://pwang.pw/ocnn.html
HIGHLIGHT:      We develop an efficient approach to impose filter orthogonality on a convolutional layer based on the doubly block-Toeplitz matrix representation of the convolutional kernel instead of using the common kernel orthogonality approach, which we show is only necessary but not sufficient for ensuring orthogonal convolutions.

26, TITLE:      SentiLR: Linguistic Knowledge Enhanced Language Representation for Sentiment Analysis
http://arxiv.org/abs/1911.02493
AUTHORS:        Pei Ke ; Haozhe Ji ; Siyang Liu ; Xiaoyan Zhu ; Minlie Huang
COMMENTS:       13 pages, a revised version
HIGHLIGHT:      To benefit the downstream tasks in sentiment analysis, we propose a novel language representation model called SentiLR, which introduces word-level linguistic knowledge including part-of-speech tag and prior sentiment polarity from SentiWordNet.

27, TITLE:      Biometric Recognition Using Deep Learning: A Survey
http://arxiv.org/abs/1912.00271
AUTHORS:        Shervin Minaee ; Amirali Abdolrashidi ; Hang Su ; Mohammed Bennamoun ; David Zhang
COMMENTS:       Under Review
HIGHLIGHT:      In this work, we provide a comprehensive survey of more than 120 promising works on biometric recognition (including face, fingerprint, iris, palmprint, ear, voice, signature, and gait recognition), which deploy deep learning models, and show their strengths and potentials in different applications. For each biometric, we first introduce the available datasets that are widely used in the literature and their characteristics.

28, TITLE:      BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning
http://arxiv.org/abs/1805.04687
AUTHORS:        Fisher Yu ; Haofeng Chen ; Xin Wang ; Wenqi Xian ; Yingying Chen ; Fangchen Liu ; Vashisht Madhavan ; Trevor Darrell
COMMENTS:       Published at IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020
HIGHLIGHT:      Researchers are usually constrained to study a small set of problems on one dataset, while real-world computer vision applications require performing tasks of various complexities. Based on this diverse dataset, we build a benchmark for heterogeneous multitask learning and study how to solve the tasks together.

29, TITLE:      Natural Language Adversarial Attacks and Defenses in Word Level
http://arxiv.org/abs/1909.06723
AUTHORS:        Xiaosen Wang ; Hao Jin ; Kun He
COMMENTS:       16 pages, 4 figures, 7 tables
HIGHLIGHT:      Natural Language Adversarial Attacks and Defenses in Word Level

30, TITLE:      Specific Single- and Multi-Objective Evolutionary Algorithms for the Chance-Constrained Knapsack Problem
http://arxiv.org/abs/2004.03205
AUTHORS:        Yue Xie ; Aneta Neumann ; Frank Neumann
COMMENTS:       Accepted for oral presentation at GECCO 2020
HIGHLIGHT:      In this paper, consider problem-specific single-objective and multi-objective approaches for the problem.

31, TITLE:      Automated Utterance Generation
http://arxiv.org/abs/2004.03484
AUTHORS:        Soham Parikh ; Quaizar Vohra ; Mitul Tiwari
COMMENTS:       AAAI/IAAI-20, Emerging Application Track
HIGHLIGHT:      In this paper, we propose an utterance generation system which 1) uses extractive summarization to extract important sentences from the description, 2) uses multiple paraphrasing techniques to generate a diverse set of paraphrases of the title and summary sentences, and 3) selects good candidate paraphrases with the help of a novel candidate selection algorithm.

32, TITLE:      OccuSeg: Occupancy-aware 3D Instance Segmentation
http://arxiv.org/abs/2003.06537
AUTHORS:        Lei Han ; Tian Zheng ; Lan Xu ; Lu Fang
COMMENTS:       CVPR 2020, video this https URL https://youtu.be/co7y6LQ7Kqc
HIGHLIGHT:      In this paper, we define "3D occupancy size", as the number of voxels occupied by each instance.

33, TITLE:      Instance-wise Depth and Motion Learning from Monocular Videos
http://arxiv.org/abs/1912.09351
AUTHORS:        Seokju Lee ; Sunghoon Im ; Stephen Lin ; In So Kweon
COMMENTS:       Project page at https://sites.google.com/site/seokjucv/home/instadm
HIGHLIGHT:      We present an end-to-end joint training framework that explicitly models 6-DoF motion of multiple dynamic objects, ego-motion and depth in a monocular camera setup without supervision.

34, TITLE:      Text Complexity Classification Based on Linguistic Information: Application to Intelligent Tutoring of ESL
http://arxiv.org/abs/2001.01863
AUTHORS:        M. Zakaria Kurdi
COMMENTS:       This is an unpublished pre-print, the JDMDH journal requires submission to arxiv.org before the submission to the journal (see the link: https://jdmdh.episciences.org/page/submissions#)
HIGHLIGHT:      The goal of this work is to build a classifier that can identify text complexity within the context of teaching reading to English as a Second Language (ESL) learners.

35, TITLE:      When, and Why, Simple Methods Fail. Lessons Learned from Hyperparameter Tuning in Software Analytics (and Elsewhere)
http://arxiv.org/abs/1912.04061
AUTHORS:        Amritanshu Agrawal ; Tim Menzies
COMMENTS:       12 Pages, In Submission to FSE 2020
HIGHLIGHT:      Our conclusion will be that we can exploit this special properties of SE to great effect; specifically, to find better optimizations for SE tasks via a tactic called "dodging" (explained in this paper).

36, TITLE:      HAMLET -- A Learning Curve-Enabled Multi-Armed Bandit for Algorithm Selection
http://arxiv.org/abs/2001.11261
AUTHORS:        Mischa Schmidt ; Julia Gastinger ; Sébastien Nicolas ; Anett Schülke
COMMENTS:       8 pages, 8 figures; Revision for IJCNN 2020: International Joint Conference on Neural Networks
HIGHLIGHT:      This work addresses that insight by introducing HAMLET, which extends the bandit approach with learning curve extrapolation and computation time-awareness for selecting among a set of machine learning algorithms.

37, 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:       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.

38, TITLE:      Super-resolution of multispectral satellite images using convolutional neural networks
http://arxiv.org/abs/2002.00580
AUTHORS:        M. U. Müller ; N. Ekhtiari ; R. M. Almeida ; C. Rieke
COMMENTS:       To be published in the ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences: https://www.isprs.org/publications/annals.aspx, proceedings of the XXIV ISPRS Congress, 14-20 June 2020, Nice, France
HIGHLIGHT:      In this paper we propose a method to train state-of-the-art CNNs using pairs of lower-resolution multispectral and high-resolution pan-sharpened image tiles in order to create super-resolved analytic images.

39, TITLE:      Universal Self-Attention Network for Graph Classification
http://arxiv.org/abs/1909.11855
AUTHORS:        Dai Quoc Nguyen ; Tu Dinh Nguyen ; Dinh Phung
COMMENTS:       Our code is available at: https://github.com/daiquocnguyen/U2GNN
HIGHLIGHT:      To this end, we present U2GNN -- a novel embedding model leveraging the strength of the transformer self-attention network -- to learn low-dimensional embeddings of graphs.

40, TITLE:      Gliding vertex on the horizontal bounding box for multi-oriented object detection
http://arxiv.org/abs/1911.09358
AUTHORS:        Yongchao Xu ; Mingtao Fu ; Qimeng Wang ; Yukang Wang ; Kai Chen ; Gui-Song Xia ; Xiang Bai
COMMENTS:       Accepted by TPAMI 2020. The experiments of pedestrian detection are updated as the benchmark has been changed
HIGHLIGHT:      In this paper, we propose a simple yet effective framework to detect multi-oriented objects.

41, TITLE:      Generalization bounds for deep convolutional neural networks
http://arxiv.org/abs/1905.12600
AUTHORS:        Philip M. Long ; Hanie Sedghi
COMMENTS:       Published as a conference paper at ICLR 2020
HIGHLIGHT:      We present experiments using CIFAR-10 with varying hyperparameters of a deep convolutional network, comparing our bounds with practical generalization gaps.

42, TITLE:      Is That a Chair? Imagining Affordances Using Simulations of an Articulated Human Body
http://arxiv.org/abs/1909.07572
AUTHORS:        Hongtao Wu ; Deven Misra ; Gregory S. Chirikjian
COMMENTS:       7 pages, 6 figures. Accepted to ICRA2020
HIGHLIGHT:      In this paper, we propose a novel method to provide robots with an ability to imagine object affordances using physical simulations.

43, TITLE:      How effective can simple ordinal peer grading be?
http://arxiv.org/abs/1602.07985
AUTHORS:        Ioannis Caragiannis ; George A. Krimpas ; Alexandros A. Voudouris
COMMENTS:       39 pages
HIGHLIGHT:      When statistical information about the grading behaviour of students is available (in terms of a noise matrix that characterizes the grading behaviour of the average student from a student population), the framework can be used to compute the optimal rule from this class with respect to a series of performance objectives that compare the ranking returned by the aggregation rule to the underlying ground truth ranking.

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

45, TITLE:      DuDoRNet: Learning a Dual-Domain Recurrent Network for Fast MRI Reconstruction with Deep T1 Prior
http://arxiv.org/abs/2001.03799
AUTHORS:        Bo Zhou ; S. Kevin Zhou
COMMENTS:       CVPR2020
HIGHLIGHT:      In this work, we address the above two limitations by proposing a Dual Domain Recurrent Network (DuDoRNet) with deep T1 prior embedded to simultaneously recover k-space and images for accelerating the acquisition of MRI with a long imaging protocol.

46, TITLE:      Reinforcement Learning in Non-Stationary Environments
http://arxiv.org/abs/1905.03970
AUTHORS:        Sindhu Padakandla ; Prabuchandran K. J ; Shalabh Bhatnagar
HIGHLIGHT:      In this paper, we thus consider the problem of developing RL methods that obtain optimal decisions in a non-stationary environment.

47, TITLE:      Sample Efficient Ensemble Learning with Catalyst.RL
http://arxiv.org/abs/2003.14210
AUTHORS:        Sergey Kolesnikov ; Valentin Khrulkov
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:1903.00027
HIGHLIGHT:      We present Catalyst.RL, an open-source PyTorch framework for reproducible and sample efficient reinforcement learning (RL) research.

48, TITLE:      Fictitious Play Outperforms Counterfactual Regret Minimization
http://arxiv.org/abs/2001.11165
AUTHORS:        Sam Ganzfried
HIGHLIGHT:      We compare the performance of two popular iterative algorithms, fictitious play and counterfactual regret minimization, in approximating Nash equilibrium in multiplayer games.

49, TITLE:      S2ORC: The Semantic Scholar Open Research Corpus
http://arxiv.org/abs/1911.02782
AUTHORS:        Kyle Lo ; Lucy Lu Wang ; Mark Neumann ; Rodney Kinney ; Dan S. Weld
COMMENTS:       14 pages, 4 figures, 6 appendices
HIGHLIGHT:      We introduce S2ORC, a large contextual citation graph of English-language academic papers from multiple scientific domains; the corpus consists of 81.1M papers, 380.5M citation edges, and associated paper metadata.

50, TITLE:      Lower Bounds for Function Inversion with Quantum Advice
http://arxiv.org/abs/1911.09176
AUTHORS:        Kai-Min Chung ; Tai-Ning Liao ; Luowen Qian
COMMENTS:       ITC full version
HIGHLIGHT:      In this work, we give the same asymptotic lower bound to fully quantum algorithms for inverting functions for fully quantum algorithms under the regime where $M = O(N)$.

51, TITLE:      Towards Stabilizing Batch Statistics in Backward Propagation of Batch Normalization
http://arxiv.org/abs/2001.06838
AUTHORS:        Junjie Yan ; Ruosi Wan ; Xiangyu Zhang ; Wei Zhang ; Yichen Wei ; Jian Sun
COMMENTS:       ICLR2020; https://github.com/megvii-model/MABN
HIGHLIGHT:      In this paper, we reveal that there are two extra batch statistics involved in backward propagation of BN, on which has never been well discussed before.

52, TITLE:      DSGN: Deep Stereo Geometry Network for 3D Object Detection
http://arxiv.org/abs/2001.03398
AUTHORS:        Yilun Chen ; Shu Liu ; Xiaoyong Shen ; Jiaya Jia
COMMENTS:       Accepted by CVPR 2020 (Camera Ready)
HIGHLIGHT:      For the first time, we provide a simple and effective one-stage stereo-based 3D detection pipeline that jointly estimates the depth and detects 3D objects in an end-to-end learning manner.

53, TITLE:      Unsupervised Cross-lingual Representation Learning at Scale
http://arxiv.org/abs/1911.02116
AUTHORS:        Alexis Conneau ; Kartikay Khandelwal ; Naman Goyal ; Vishrav Chaudhary ; Guillaume Wenzek ; Francisco Guzmán ; Edouard Grave ; Myle Ott ; Luke Zettlemoyer ; Veselin Stoyanov
COMMENTS:       ACL 2020 (+ updated results)
HIGHLIGHT:      We train a Transformer-based masked language model on one hundred languages, using more than two terabytes of filtered CommonCrawl data.

54, TITLE:      Image Segmentation Using Deep Learning: A Survey
http://arxiv.org/abs/2001.05566
AUTHORS:        Shervin Minaee ; Yuri Boykov ; Fatih Porikli ; Antonio Plaza ; Nasser Kehtarnavaz ; Demetri Terzopoulos
HIGHLIGHT:      In this survey, we provide a comprehensive review of the literature at the time of this writing, covering a broad spectrum of pioneering works for semantic and instance-level segmentation, including fully convolutional pixel-labeling networks, encoder-decoder architectures, multi-scale and pyramid based approaches, recurrent networks, visual attention models, and generative models in adversarial settings.

55, TITLE:      Learning Deep Conditional Target Densities for Accurate Regression
http://arxiv.org/abs/1909.12297
AUTHORS:        Fredrik K. Gustafsson ; Martin Danelljan ; Goutam Bhat ; Thomas B. Schön
HIGHLIGHT:      We address these issues by proposing a general and conceptually simple regression method with a clear probabilistic interpretation.

56, TITLE:      Knife and Threat Detectors
http://arxiv.org/abs/2004.03366
AUTHORS:        David A. Noever ; Sam E. Miller Noever
HIGHLIGHT:      We present three complementary methods for scoring automated threat identification using multiple knife image datasets, each with the goal of narrowing down possible assault intentions while minimizing misidentifying false positives and risky false negatives.

57, TITLE:      Deep learning approaches in food recognition
http://arxiv.org/abs/2004.03357
AUTHORS:        Chairi Kiourt ; George Pavlidis ; Stella Markantonatou
COMMENTS:       26 pages, 10 figures, book chapter for Machine Learning Paradigms - Advances in Theory and Applications of Deep Learning
HIGHLIGHT:      Traditional image analysis approaches have achieved low classification accuracy in the past, whereas deep learning approaches enabled the identification of food types and their ingredients.

58, TITLE:      Attribution in Scale and Space
http://arxiv.org/abs/2004.03383
AUTHORS:        Shawn Xu ; Subhashini Venugopalan ; Mukund Sundararajan
COMMENTS:       CVPR 2020 camera-ready. Code is available at https://github.com/PAIR-code/saliency
HIGHLIGHT:      We propose a new technique called \emph{Blur Integrated Gradients}.