2020.06.30.txt

==========New Papers==========
1, TITLE:       A Tool for Automatic Estimation of Patient Position in Spinal CT Data
http://arxiv.org/abs/2006.15330
AUTHORS:        Roman Jakubicek ; Tomas Vicar ; Jiri Chmelik
COMMENTS:       6 pages, 3 figures, submitted on EMBEC 2020; the paper has not been reviewed yet
HIGHLIGHT:      This paper presents a tool for automatic rotation of CT data into a standardized (HFS) patient position.

2, TITLE:       Compositional Video Synthesis with Action Graphs
http://arxiv.org/abs/2006.15327
AUTHORS:        Amir Bar ; Roei Herzig ; Xiaolong Wang ; Gal Chechik ; Trevor Darrell ; Amir Globerson
HIGHLIGHT:      Towards this end, we introduce a generative model (AG2Vid) based on Action Graphs, a natural and convenient structure that represents the dynamics of actions between objects over time.

3, TITLE:       Chroma Intra Prediction with attention-based CNN architectures
http://arxiv.org/abs/2006.15349
AUTHORS:        Marc Górriz ; Saverio Blasi ; Alan F. Smeaton ; Noel E. O'Connor ; Marta Mrak
COMMENTS:       27th IEEE International Conference on Image Processing, 25-28 Oct 2020, Abu Dhabi, United Arab Emirates
HIGHLIGHT:      This paper proposes a new neural network architecture for cross-component intra-prediction.

4, TITLE:       End-Effect Exploration Drive for Effective Motor Learning
http://arxiv.org/abs/2006.15960
AUTHORS:        Emmanuel Daucé
COMMENTS:       6 pages, 3 figures, submitted to IWAI 2020 (1st International Workshop on Active Inference)
HIGHLIGHT:      End-Effect Exploration Drive for Effective Motor Learning

5, TITLE:       Forgery Detection in a Questioned Hyperspectral Document Image using K-means Clustering
http://arxiv.org/abs/2006.16057
AUTHORS:        Maria Yaseen ; Rammal Aftab Ahmed ; Rimsha Mahrukh
COMMENTS:       5 pages,6 figures
HIGHLIGHT:      In this paper an extensive ink mismatch detection technique is presented which uses KMean Clustering to identify different inks on the basis of their unique spectral response and separates them into different clusters.

6, TITLE:       The Evolutionary Dynamics of Independent Learning Agents in Population Games
http://arxiv.org/abs/2006.16068
AUTHORS:        Shuyue Hu ; Chin-Wing Leung ; Ho-fung Leung ; Harold Soh
HIGHLIGHT:      This paper presents a formal relation between stochastic processes and the dynamics of independent learning agents who reason based on the reward signals.

7, TITLE:       Patch SVDD: Patch-level SVDD for Anomaly Detection and Segmentation
http://arxiv.org/abs/2006.16067
AUTHORS:        Jihun Yi ; Sungroh Yoon
HIGHLIGHT:      In this paper, we tackle the problem of image anomaly detection and segmentation.

8, TITLE:       AerialMPTNet: Multi-Pedestrian Tracking in Aerial Imagery Using Temporal and Graphical Features
http://arxiv.org/abs/2006.15457
AUTHORS:        Maximilian Kraus ; Seyed Majid Azimi ; Emec Ercelik ; Reza Bahmanyar ; Peter Reinartz ; Alois Knoll
COMMENTS:       ICPR 2020
HIGHLIGHT:      In this paper, we propose AerialMPTNet, a novel approach for multi-pedestrian tracking in geo-referenced aerial imagery by fusing appearance features from a Siamese Neural Network, movement predictions from a Long Short-Term Memory, and pedestrian interconnections from a GraphCNN.

9, TITLE:       Measuring Memorization Effect in Word-Level Neural Networks Probing
http://arxiv.org/abs/2006.16082
AUTHORS:        Rudolf Rosa ; Tomáš Musil ; David Mareček
COMMENTS:       Accepted to TSD 2020. Will be published in Springer LNCS
HIGHLIGHT:      In our work, we propose a simple general method for measuring the memorization effect, based on a symmetric selection of comparable sets of test words seen versus unseen in training.

10, TITLE:      A Deep Reinforced Model for Zero-Shot Cross-Lingual Summarization with Bilingual Semantic Similarity Rewards
http://arxiv.org/abs/2006.15454
AUTHORS:        Zi-Yi Dou ; Sachin Kumar ; Yulia Tsvetkov
HIGHLIGHT:      In this work, we propose an end-to-end cross-lingual text summarization model.

11, TITLE:      Layered Stereo by Cooperative Grouping with Occlusion
http://arxiv.org/abs/2006.16094
AUTHORS:        Jialiang Wang ; Todd Zickler
HIGHLIGHT:      In this work, we introduce a layered approach to stereo that explicitly incorporates occlusions.

12, TITLE:      Frequency learning for image classification
http://arxiv.org/abs/2006.15476
AUTHORS:        José Augusto Stuchi ; Levy Boccato ; Romis Attux
HIGHLIGHT:      In this context, this paper presents a new approach for exploring the Fourier transform of the input images, which is composed of trainable frequency filters that boost discriminative components in the spectrum.

13, TITLE:      MvMM-RegNet: A new image registration framework based on multivariate mixture model and neural network estimation
http://arxiv.org/abs/2006.15573
AUTHORS:        Xinzhe Luo ; Xiahai Zhuang
COMMENTS:       Accepted for publication at MICCAI 2020
HIGHLIGHT:      In this paper, we propose a new image registration framework based on multivariate mixture model (MvMM) and neural network estimation.

14, TITLE:      Multi-Person Pose Regression via Pose Filtering and Scoring
http://arxiv.org/abs/2006.15576
AUTHORS:        Huixin Miao ; Junqi Lin ; Junjie Cao
COMMENTS:       8 pages, 4 figures
HIGHLIGHT:      In this work, we propose an end-to-end network framework for multi-person pose regression to predict the instance-aware keypoints directly.

15, TITLE:      Fabric Image Representation Encoding Networks for Large-scale 3D Medical Image Analysis
http://arxiv.org/abs/2006.15578
AUTHORS:        Siyu Liu ; Wei Dai ; Craig Engstrom ; Jurgen Fripp ; Peter B. Greer ; Stuart Crozier ; Jason A. Dowling ; Shekhar S Chandra
COMMENTS:       12 pages, 10 figures
HIGHLIGHT:      In this work, a novel 3D segmentation network, Fabric Image Representation Networks (FIRENet), is proposed to extract and encode generalisable feature representations from multiple medical image datasets in a large-scale manner.

16, TITLE:      Self-Attention Networks for Intent Detection
http://arxiv.org/abs/2006.15585
AUTHORS:        Sevinj Yolchuyeva ; Géza Németh ; Bálint Gyires-Tóth
COMMENTS:       Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2019)
HIGHLIGHT:      In this paper, we present a novel intent detection system which is based on a self-attention network and a Bi-LSTM.

17, TITLE:      BERTology Meets Biology: Interpreting Attention in Protein Language Models
http://arxiv.org/abs/2006.15222
AUTHORS:        Jesse Vig ; Ali Madani ; Lav R. Varshney ; Caiming Xiong ; Richard Socher ; Nazneen Fatema Rajani
HIGHLIGHT:      Through the lens of attention, we analyze the inner workings of the Transformer and explore how the model discerns structural and functional properties of proteins.

18, TITLE:      Perception-Prediction-Reaction Agents for Deep Reinforcement Learning
http://arxiv.org/abs/2006.15223
AUTHORS:        Adam Stooke ; Valentin Dalibard ; Siddhant M. Jayakumar ; Wojciech M. Czarnecki ; Max Jaderberg
HIGHLIGHT:      We introduce a new recurrent agent architecture and associated auxiliary losses which improve reinforcement learning in partially observable tasks requiring long-term memory.

19, TITLE:      A lateral semicircular canal segmentation based geometric calibration for human temporal bone CT Image
http://arxiv.org/abs/2006.15588
AUTHORS:        Xiaoguang Li ; Peng Fu ; Hongxia Yin ; ZhenChang Wang ; Li Zhuo ; Hui Zhang
HIGHLIGHT:      We propose an automatic calibration algorithm for temporal bone CT images.

20, TITLE:      On the Relationship Between Probabilistic Circuits and Determinantal Point Processes
http://arxiv.org/abs/2006.15233
AUTHORS:        Honghua Zhang ; Steven Holtzen ; Guy Van den Broeck
HIGHLIGHT:      We propose a unified analysis and shared language for discussing DPPs and PCs.

21, TITLE:      Computation Offloading in Multi-Access Edge Computing Networks: A Multi-Task Learning Approach
http://arxiv.org/abs/2006.16104
AUTHORS:        Bo Yang ; Xuelin Cao ; Joshua Bassey ; Xiangfang Li ; Timothy Kroecker ; Lijun Qian
HIGHLIGHT:      In this paper, we propose a dynamic offloading framework for the MEC network, in which the uplink non-orthogonal multiple access (NOMA) is used to enable multiple devices to upload their tasks via the same frequency band.

22, TITLE:      Iris Recognition: Inherent Binomial Degrees of Freedom
http://arxiv.org/abs/2006.16107
AUTHORS:        J. Michael Rozmus
COMMENTS:       4 pages, 6 figures
HIGHLIGHT:      This paper shows by direct pixel-by-pixel comparison of high-quality iris images that the inherent number of degrees of freedom embodied in the human iris, independent of any encoding, is at least 536.

23, TITLE:      GramGAN: Deep 3D Texture Synthesis From 2D Exemplars
http://arxiv.org/abs/2006.16112
AUTHORS:        Tiziano Portenier ; Siavash Bigdeli ; Orçun Göksel
HIGHLIGHT:      We present a novel texture synthesis framework, enabling the generation of infinite, high-quality 3D textures given a 2D exemplar image.

24, TITLE:      Attention-Guided Generative Adversarial Network to Address Atypical Anatomy in Modality Transfer
http://arxiv.org/abs/2006.15264
AUTHORS:        Hajar Emami ; Ming Dong ; Carri K. Glide-Hurst
HIGHLIGHT:      In this paper, we propose a novel spatial attention-guided generative adversarial network (attention-GAN) model to generate accurate synCTs using T1-weighted MRI images as the input to address atypical anatomy.

25, TITLE:      PCLNet: A Practical Way for Unsupervised Deep PolSAR Representations and Few-Shot Classification
http://arxiv.org/abs/2006.15351
AUTHORS:        Lamei Zhang ; Siyu Zhang ; Bin Zou ; Hongwei Dong
COMMENTS:       16 pages, 16 figures
HIGHLIGHT:      To handle this problem, in this paper, learning transferrable representations from unlabeled PolSAR data through convolutional architectures is explored for the first time.

26, TITLE:      MiniNet: An extremely lightweight convolutional neural network for real-time unsupervised monocular depth estimation
http://arxiv.org/abs/2006.15350
AUTHORS:        Jun Liu ; Qing Li ; Rui Cao ; Wenming Tang ; Guoping Qiu
HIGHLIGHT:      To address this issue, we propose a new powerful network with a recurrent module to achieve the capability of a deep network while at the same time maintaining an extremely lightweight size for real-time high performance unsupervised monocular depth prediction from video sequences.

27, TITLE:      Coloured noise time series as appropriate models for environmental variation in artificial evolutionary systems
http://arxiv.org/abs/2006.16204
AUTHORS:        Matt Grove ; James M. Borg ; Fiona Polack
COMMENTS:       8 pages, 4 figures, 2020 conference on Artificial Life
HIGHLIGHT:      Ultimately we argue that Artificial Life as a field should embrace the use of coloured noise to produce models of environmental variability.

28, TITLE:      An Evoked Potential-Guided Deep Learning Brain Representation For Visual Classification
http://arxiv.org/abs/2006.15357
AUTHORS:        Xianglin Zheng ; Zehong Cao ; Quan Bai
COMMENTS:       This paper is submitting to ICONIP 2020
HIGHLIGHT:      In this study, we proposed a deep learning framework guided by the visual evoked potentials, called the Event-Related Potential (ERP)-Long short-term memory (LSTM) framework, extracted by EEG signals for visual classification.

29, TITLE:      Towards the Study of Morphological Processing of the Tangkhul Language
http://arxiv.org/abs/2006.16212
AUTHORS:        Mirinso Shadang ; Navanath Saharia ; Thoudam Doren Singh
COMMENTS:       In proceeding of Regional International Conference on Natural Language Processing (regICON) 2017, 3rd and 4th November 2017, Imphal, India
HIGHLIGHT:      We use a small corpus collected from different sources of text books, short stories and articles of other topics.

30, TITLE:      ReMarNet: Conjoint Relation and Margin Learning for Small-Sample Image Classification
http://arxiv.org/abs/2006.15366
AUTHORS:        Xiaoxu Li ; Liyun Yu ; Xiaochen Yang ; Zhanyu Ma ; Jing-Hao Xue ; Jie Cao ; Jun Guo
COMMENTS:       IEEE TCSVT 2020
HIGHLIGHT:      In this paper, we propose to enhance the discriminative power of features from a new perspective by introducing a novel neural network termed Relation-and-Margin learning Network (ReMarNet).

31, TITLE:      MTStereo 2.0: improved accuracy of stereo depth estimation withMax-trees
http://arxiv.org/abs/2006.15373
AUTHORS:        Rafael Brandt ; Nicola Strisciuglio ; Nicolai Petkov
HIGHLIGHT:      In this paper, we propose a stereo matching method, called MTStereo 2.0, for limited-resource systems that require efficient and accurate depth estimation.

32, TITLE:      Liquid Resource Types
http://arxiv.org/abs/2006.16233
AUTHORS:        Tristan Knoth ; Di Wang ; Adam Reynolds ; Jan Hoffmann ; Nadia Polikarpova
HIGHLIGHT:      This article presents liquid resource types, a technique for automatically verifying the resource consumption of functional programs.

33, TITLE:      The Generalized Independent and Dominating Set Problems on Unit Disk Graphs
http://arxiv.org/abs/2006.15381
AUTHORS:        Sangram K. Jena ; Ramesh K. Jallu ; Gautam K. Das ; Subhas C. Nandy
HIGHLIGHT:      In this article, we study a generalized version of the maximum independent set and minimum dominating set problems, namely, the maximum $d$-distance independent set problem and the minimum $d$-distance dominating set problem on unit disk graphs for a positive integer $d>0$.

34, TITLE:      Self-Supervised MultiModal Versatile Networks
http://arxiv.org/abs/2006.16228
AUTHORS:        Jean-Baptiste Alayrac ; Adrià Recasens ; Rosalia Schneider ; Relja Arandjelović ; Jason Ramapuram ; Jeffrey De Fauw ; Lucas Smaira ; Sander Dieleman ; Andrew Zisserman
HIGHLIGHT:      In this work, we learn representations using self-supervision by leveraging three modalities naturally present in videos: vision, audio and language.

35, TITLE:      Light Pose Calibration for Camera-light Vision Systems
http://arxiv.org/abs/2006.15389
AUTHORS:        Yifan Song ; Furkan Elibol ; Mengkun She ; David Nakath ; Kevin Köser
HIGHLIGHT:      This paper presents a novel light calibration approach by taking multi-view and -distance images of a reference plane in order to provide pose information of the employed light sources to the computer vision system.

36, TITLE:      Automated Stitching of Coral Reef Images and Extraction of Features for Damselfish Shoaling Behavior Analysis
http://arxiv.org/abs/2006.15478
AUTHORS:        Riza Rae Pineda ; Kristofer delas Peñas ; Dana Manogan
HIGHLIGHT:      To effectively analyze shoaling behavior given the issues posed by capturing data in the wild, we propose a pre-processing system that utilizes color correction and image stitching techniques and extracts behavior features for manual analysis.

37, TITLE:      Interpretable Deepfake Detection via Dynamic Prototypes
http://arxiv.org/abs/2006.15473
AUTHORS:        Loc Trinh ; Michael Tsang ; Sirisha Rambhatla ; Yan Liu
HIGHLIGHT:      To this end, we propose Dynamic Prototype Network (DPNet) - a simple, interpretable, yet effective solution that leverages dynamic representations (i.e., prototypes) to explain deepfake visual dynamics.

38, TITLE:      Deep Sea Robotic Imaging Simulator for UUV Development
http://arxiv.org/abs/2006.15398
AUTHORS:        Yifan Song ; David Nakath ; Mengkun She ; Furkan Elibol ; Kevin Köser
HIGHLIGHT:      Thus this paper presents a physical model-based image simulation solution which uses in-air and depth information as inputs to generate underwater images for robotics in deep ocean scenarios.

39, TITLE:      Laplacian Regularized Few-Shot Learning
http://arxiv.org/abs/2006.15486
AUTHORS:        Imtiaz Masud Ziko ; Jose Dolz ; Eric Granger ; Ismail Ben Ayed
COMMENTS:       ICML 2020 paper
HIGHLIGHT:      We propose a transductive Laplacian-regularized inference for few-shot tasks.

40, TITLE:      Bottom-Up Human Pose Estimation by Ranking Heatmap-Guided Adaptive Keypoint Estimates
http://arxiv.org/abs/2006.15480
AUTHORS:        Ke Sun ; Zigang Geng ; Depu Meng ; Bin Xiao ; Dong Liu ; Zhaoxiang Zhang ; Jingdong Wang
HIGHLIGHT:      The typical bottom-up human pose estimation framework includes two stages, keypoint detection and grouping.

41, TITLE:      Video Representation Learning with Visual Tempo Consistency
http://arxiv.org/abs/2006.15489
AUTHORS:        Ceyuan Yang ; Yinghao Xu ; Bo Dai ; Bolei Zhou
COMMENTS:       Technical report. Models are available at https://github.com/decisionforce/VTHCL
HIGHLIGHT:      In this work, we demonstrate that visual tempo can also serve as a self-supervision signal for video representation learning.

42, TITLE:      Rethinking the Positional Encoding in Language Pre-training
http://arxiv.org/abs/2006.15595
AUTHORS:        Guolin Ke ; Di He ; Tie-Yan Liu
HIGHLIGHT:      In this work, we investigate the problems in the previous formulations and propose a new positional encoding method for BERT called Transformer with Untied Positional Encoding (TUPE).

43, TITLE:      A Framework for Pre-processing of Social Media Feeds based on Integrated Local Knowledge Base
http://arxiv.org/abs/2006.15854
AUTHORS:        Taiwo Kolajo ; Olawande Daramola ; Ayodele Adebiyi ; Seth Aaditeshwar
COMMENTS:       38 pages, 5 figures, 6 tables
HIGHLIGHT:      This paper proposes an improved framework for pre-processing of social media feeds for better performance.

44, TITLE:      MoNet3D: Towards Accurate Monocular 3D Object Localization in Real Time
http://arxiv.org/abs/2006.16007
AUTHORS:        Xichuan Zhou ; Yicong Peng ; Chunqiao Long ; Fengbo Ren ; Cong Shi
HIGHLIGHT:      The MoNet3D method incorporates prior knowledge of the spatial geometric correlation of neighbouring objects into the deep neural network training process to improve the accuracy of 3D object localization.

45, TITLE:      OpenDVC: An Open Source Implementation of the DVC Video Compression Method
http://arxiv.org/abs/2006.15862
AUTHORS:        Ren Yang ; Luc Van Gool ; Radu Timofte
COMMENTS:       Technical report of OpenDVC; the project page is at https://github.com/RenYang-home/OpenDVC
HIGHLIGHT:      We introduce an open source Tensorflow implementation of the Deep Video Compression (DVC) method in this technical report.

46, TITLE:      Abnormal activity capture from passenger flow of elevator based on unsupervised learning and fine-grained multi-label recognition
http://arxiv.org/abs/2006.15873
AUTHORS:        Chunhua Jia ; Wenhai Yi ; Yu Wu ; Hui Huang ; Lei Zhang ; Leilei Wu
COMMENTS:       9 pages, 8 figures, submitted to 34th Conference on Neural Information Processing System(NeurIPS 2020)
HIGHLIGHT:      More specifically in our implementation we propose GraftNet, a solution for fine-grained multi-label recognition task, to recognize human attributes, e.g. gender, age, appearance, and occupation.

47, TITLE:      Deep Involutive Generative Models for Neural MCMC
http://arxiv.org/abs/2006.15167
AUTHORS:        Span Spanbauer ; Cameron Freer ; Vikash Mansinghka
COMMENTS:       13 pages, 6 figures
HIGHLIGHT:      We introduce deep involutive generative models, a new architecture for deep generative modeling, and use them to define Involutive Neural MCMC, a new approach to fast neural MCMC.

48, TITLE:      Propagation for Dynamic Continuous Time Chain Event Graphs
http://arxiv.org/abs/2006.15865
AUTHORS:        Aditi Shenvi ; Jim Q. Smith
HIGHLIGHT:      We present a tractable exact inferential scheme analogous to the scheme in Kj{\ae}rulff (1992) for discrete Dynamic Bayesian Networks (DBNs) which employs standard junction tree inference by "unrolling" the DBN.

49, TITLE:      Intrinsic Autoencoders for Joint Neural Rendering and Intrinsic Image Decomposition
http://arxiv.org/abs/2006.16011
AUTHORS:        Hassan Abu Alhaija ; Siva Karthik Mustikovela ; Justus Thies ; Matthias Nießner ; Andreas Geiger ; Carsten Rother
HIGHLIGHT:      The main contribution of this work is to lift this restriction by training a neural rendering algorithm from unpaired data.

50, TITLE:      Compressive MR Fingerprinting reconstruction with Neural Proximal Gradient iterations
http://arxiv.org/abs/2006.15271
AUTHORS:        Dongdong Chen ; Mike E. Davies ; Mohammad Golbabaee
COMMENTS:       To appear in MICCAI 2020
HIGHLIGHT:      To address this, we propose ProxNet, a learned proximal gradient descent framework that directly incorporates the forward acquisition and Bloch dynamic models within a recurrent learning mechanism.

51, TITLE:      Explainable 3D Convolutional Neural Networks by Learning Temporal Transformations
http://arxiv.org/abs/2006.15983
AUTHORS:        Gabriëlle Ras ; Luca Ambrogioni ; Pim Haselager ; Marcel A. J. van Gerven ; Umut Güçlü
COMMENTS:       10 pages, 5 figures, 4 tables
HIGHLIGHT:      In this paper we introduce the temporally factorized 3D convolution (3TConv) as an interpretable alternative to the regular 3D convolution (3DConv).

52, TITLE:      1st Place Solution for Waymo Open Dataset Challenge -- 3D Detection and Domain Adaptation
http://arxiv.org/abs/2006.15505
AUTHORS:        Zhuangzhuang Ding ; Yihan Hu ; Runzhou Ge ; Li Huang ; Sijia Chen ; Yu Wang ; Jie Liao
HIGHLIGHT:      In this technical report, we introduce our winning solution "HorizonLiDAR3D" for the 3D detection track and the domain adaptation track in Waymo Open Dataset Challenge at CVPR 2020.

53, TITLE:      1st Place Solutions for Waymo Open Dataset Challenges -- 2D and 3D Tracking
http://arxiv.org/abs/2006.15506
AUTHORS:        Yu Wang ; Sijia Chen ; Li Huang ; Runzhou Ge ; Yihan Hu ; Zhuangzhuang Ding ; Jie Liao
HIGHLIGHT:      This technical report presents the online and real-time 2D and 3D multi-object tracking (MOT) algorithms that reached the 1st places on both Waymo Open Dataset 2D tracking and 3D tracking challenges.

54, TITLE:      2nd Place Solution for Waymo Open Dataset Challenge -- 2D Object Detection
http://arxiv.org/abs/2006.15507
AUTHORS:        Sijia Chen ; Yu Wang ; Li Huang ; Runzhou Ge ; Yihan Hu ; Zhuangzhuang Ding ; Jie Liao
HIGHLIGHT:      In this report, we introduce a state-of-the-art 2D object detection system for autonomous driving scenarios.

55, TITLE:      Shape from Projections via Differentiable Forward Projector
http://arxiv.org/abs/2006.16120
AUTHORS:        Jakeoung Koo ; Qiongyang Chen ; J. Andreas Bærentzen ; Anders B. Dahl ; Vedrana A. Dahl
HIGHLIGHT:      In this paper, we propose a differentiable forward projector for 3D meshes, to bridge the gap between the forward model for 3D surfaces and optimization.

56, TITLE:      Human Activity Recognition based on Dynamic Spatio-Temporal Relations
http://arxiv.org/abs/2006.16132
AUTHORS:        Zhenyu Liu ; Yaqiang Yao ; Yan Liu ; Yuening Zhu ; Zhenchao Tao ; Lei Wang ; Yuhong Feng
HIGHLIGHT:      In this paper, we develop a method for human activity recognition that tackles these two issues.

57, TITLE:      Layer Sparsity in Neural Networks
http://arxiv.org/abs/2006.15604
AUTHORS:        Mohamed Hebiri ; Johannes Lederer
HIGHLIGHT:      In this paper, we discuss sparsity in the framework of neural networks.

58, TITLE:      Localization Uncertainty Estimation for Anchor-Free Object Detection
http://arxiv.org/abs/2006.15607
AUTHORS:        Youngwan Lee ; Joong-won Hwang ; Hyung-Il Kim ; Kimin Yun ; Joungyoul Park
HIGHLIGHT:      Therefore, we propose a new object detector called Gaussian-FCOS that estimates the localization uncertainty based on an anchor-free detector that captures the uncertainty of similar property with four directions of box offsets (left, right, top, bottom) and avoids the anchor tuning.

59, TITLE:      Shadow Removal by a Lightness-Guided Network with Training on Unpaired Data
http://arxiv.org/abs/2006.15617
AUTHORS:        Zhihao Liu ; Hui Yin ; Yang Mi ; Mengyang Pu ; Song Wang
COMMENTS:       Submitted to IEEE TIP
HIGHLIGHT:      In this paper, we present a new Lightness-Guided Shadow Removal Network (LG-ShadowNet) for shadow removal by training on unpaired data.

60, TITLE:      Analogical Image Translation for Fog Generation
http://arxiv.org/abs/2006.15618
AUTHORS:        Rui Gong ; Dengxin Dai ; Yuhua Chen ; Wen Li ; Luc Van Gool
COMMENTS:       18 pages, 9 figures, 7 tables
HIGHLIGHT:      In this work, we are interested in adding adverse weather effects, more specifically fog effects, to images taken in clear weather.

61, TITLE:      Offline Handwritten Chinese Text Recognition with Convolutional Neural Networks
http://arxiv.org/abs/2006.15619
AUTHORS:        Brian Liu ; Xianchao Xu ; Yu Zhang
COMMENTS:       6 pages, 5 figures, and 3 tables
HIGHLIGHT:      In this paper, we build the models using only the convolutional neural networks and use CTC as the loss function.

62, TITLE:      Active Finite Reward Automaton Inference and Reinforcement Learning Using Queries and Counterexamples
http://arxiv.org/abs/2006.15714
AUTHORS:        Zhe Xu ; Bo Wu ; Daniel Neider ; Ufuk Topcu
HIGHLIGHT:      We propose an active learning approach that iteratively infers finite reward automata and performs RL (specifically, q-learning) based on the inferred finite reward automata.

63, TITLE:      Improving VQA and its Explanations \\ by Comparing Competing Explanations
http://arxiv.org/abs/2006.15631
AUTHORS:        Jialin Wu ; Liyan Chen ; Raymond J. Mooney
HIGHLIGHT:      To address this issue, we present a novel framework that uses explanations for competing answers to help VQA systems select the correct answer.

64, TITLE:      Motion Pyramid Networks for Accurate and Efficient Cardiac Motion Estimation
http://arxiv.org/abs/2006.15710
AUTHORS:        Hanchao Yu ; Xiao Chen ; Humphrey Shi ; Terrence Chen ; Thomas S. Huang ; Shanhui Sun
COMMENTS:       Accepted by MICCAI2020
HIGHLIGHT:      In this paper, we propose Motion Pyramid Networks, a novel deep learning-based approach for accurate and efficient cardiac motion estimation.

65, TITLE:      Want to Identify, Extract and Normalize Adverse Drug Reactions in Tweets? Use RoBERTa
http://arxiv.org/abs/2006.16146
AUTHORS:        Katikapalli Subramanyam Kalyan ; S. Sangeetha
COMMENTS:       4 pages
HIGHLIGHT:      This paper presents our approach for task 2 and task 3 of Social Media Mining for Health (SMM4H) 2020 shared tasks.

66, TITLE:      Large Deformation Diffeomorphic Image Registration with Laplacian Pyramid Networks
http://arxiv.org/abs/2006.16148
AUTHORS:        Tony C. W. Mok ; Albert C. S. Chung
COMMENTS:       Paper accepted by MICCAI 2020
HIGHLIGHT:      In this paper, we propose a deep Laplacian Pyramid Image Registration Network, which can solve the image registration optimization problem in a coarse-to-fine fashion within the space of diffeomorphic maps.

67, TITLE:      Simulation and Optimisation of Air Conditioning Systems using Machine Learning
http://arxiv.org/abs/2006.15296
AUTHORS:        Rakshitha Godahewa ; Chang Deng ; Arnaud Prouzeau ; Christoph Bergmeir
HIGHLIGHT:      We introduce a deep-learning model based on Recurrent Neural Networks (RNN) that can predict the temperatures of a future period directly where a particular room is unoccupied and by using these predicted temperatures, we define the optimal thermal setpoints to be used inside the room during the unoccupied period.

68, TITLE:      The Heterogeneity Hypothesis: Finding Layer-Wise Dissimilated Network Architecture
http://arxiv.org/abs/2006.16242
AUTHORS:        Yawei Li ; Wen Li ; Martin Danelljan ; Kai Zhang ; Shuhang Gu ; Luc Van Gool ; Radu Timofte
COMMENTS:       Code will be available at https://github.com/ofsoundof/Heterogeneity_Hypothesis
HIGHLIGHT:      In this paper, we tackle the problem of convolutional neural network design.

69, TITLE:      The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization
http://arxiv.org/abs/2006.16241
AUTHORS:        Dan Hendrycks ; Steven Basart ; Norman Mu ; Saurav Kadavath ; Frank Wang ; Evan Dorundo ; Rahul Desai ; Tyler Zhu ; Samyak Parajuli ; Mike Guo ; Dawn Song ; Jacob Steinhardt ; Justin Gilmer
COMMENTS:       Datasets, code, and models available at https://github.com/hendrycks/imagenet-r
HIGHLIGHT:      We introduce three new robustness benchmarks consisting of naturally occurring distribution changes in image style, geographic location, camera operation, and more.

70, TITLE:      Leveraging Subword Embeddings for Multinational Address Parsing
http://arxiv.org/abs/2006.16152
AUTHORS:        Marouane Yassine ; David Beauchemin ; François Laviolette ; Luc Lamontagne
COMMENTS:       submitted to IEEE CiSt'20
HIGHLIGHT:      We propose an approach in which we employ subword embeddings and a Recurrent Neural Network architecture to build a single model capable of learning to parse addresses from multiple countries at the same time while taking into account the difference in languages and address formatting systems.

71, TITLE:      Progressive Generation of Long Text
http://arxiv.org/abs/2006.15720
AUTHORS:        Bowen Tan ; Zichao Yang ; Maruan AI-Shedivat ; Eric P. Xing ; Zhiting Hu
COMMENTS:       Code available at https://github.com/tanyuqian/progressive-generation
HIGHLIGHT:      To overcome the limitation, we propose a simple but effective method of generating text in a progressive manner, inspired by generating images from low to high resolution.

72, TITLE:      Unsupervised Learning of Video Representations via Dense Trajectory Clustering
http://arxiv.org/abs/2006.15731
AUTHORS:        Pavel Tokmakov ; Martial Hebert ; Cordelia Schmid
HIGHLIGHT:      This paper addresses the task of unsupervised learning of representations for action recognition in videos.

73, TITLE:      Roweisposes, Including Eigenposes, Supervised Eigenposes, and Fisherposes, for 3D Action Recognition
http://arxiv.org/abs/2006.15736
AUTHORS:        Benyamin Ghojogh ; Fakhri Karray ; Mark Crowley
COMMENTS:       key-words: Roweisposes, Roweis discriminant analysis, Fisherposes, eigenposes, supervised eigenposes, action recognition
HIGHLIGHT:      In this paper, we propose Roweisposes which uses Roweis discriminant analysis for generalized subspace learning.

74, TITLE:      Mapping Topic Evolution Across Poetic Traditions
http://arxiv.org/abs/2006.15732
AUTHORS:        Petr Plechac ; Thomas N. Haider
HIGHLIGHT:      Poetic traditions across languages evolved differently, but we find that certain semantic topics occur in several of them, albeit sometimes with temporal delay, or with diverging trajectories over time.

75, TITLE:      EmotionNet Nano: An Efficient Deep Convolutional Neural Network Design for Real-time Facial Expression Recognition
http://arxiv.org/abs/2006.15759
AUTHORS:        James Ren Hou Lee ; Linda Wang ; Alexander Wong
COMMENTS:       9 pages
HIGHLIGHT:      Motivated by this need for a compact, low latency, yet accurate system capable of performing FEC in real-time on low-cost embedded devices, this study proposes EmotionNet Nano, an efficient deep convolutional neural network created through a human-machine collaborative design strategy, where human experience is combined with machine meticulousness and speed in order to craft a deep neural network design catered towards real-time embedded usage.

76, TITLE:      Exploring Optimal Control With Observations at a Cost
http://arxiv.org/abs/2006.15757
AUTHORS:        Rui Aguiar ; Nikka Mofid ; Hyunji Alex Nam
COMMENTS:       8 pages, 10 figures
HIGHLIGHT:      Our approach models this problem using OpenAI Gym's Mountain Car and aims to address when to observe the patient's physiological state and partly how to intervene, as we have assumed we can only act after following an observation.

77, TITLE:      Confidence-rich grid mapping
http://arxiv.org/abs/2006.15754
AUTHORS:        Ali-akbar Agha-mohammadi ; Eric Heiden ; Karol Hausman ; Gaurav S. Sukhatme
COMMENTS:       Published at International Journal of Robotics Research (IJRR) 2019 (https://journals.sagepub.com/doi/10.1177/0278364919839762)
HIGHLIGHT:      In this work, we present confidence-rich mapping (CRM), a new algorithm for spatial grid-based mapping of the 3D environment.

78, TITLE:      Empirically Verifying Hypotheses Using Reinforcement Learning
http://arxiv.org/abs/2006.15762
AUTHORS:        Kenneth Marino ; Rob Fergus ; Arthur Szlam ; Abhinav Gupta
HIGHLIGHT:      Specifically, we aim to build an agent that, given a hypothesis about the dynamics of the world, can take actions to generate observations which can help predict whether the hypothesis is true or false.

79, TITLE:      Active Ensemble Deep Learning for Polarimetric Synthetic Aperture Radar Image Classification
http://arxiv.org/abs/2006.15771
AUTHORS:        Sheng-Jie Liu ; Haowen Luo ; Qian Shi
COMMENTS:       Accepted by GRSL
HIGHLIGHT:      In this letter, we take the advantage of active learning and propose active ensemble deep learning (AEDL) for PolSAR image classification.

80, TITLE:      Mining Persistent Activity in Continually Evolving Networks
http://arxiv.org/abs/2006.15410
AUTHORS:        Caleb Belth ; Xinyi Zheng ; Danai Koutra
COMMENTS:       9 pages, plus 2 pages of supplementary material. Accepted at KDD 2020
HIGHLIGHT:      In this work, we propose the problem of mining activity that persists through time in continually evolving networks---i.e., activity that repeatedly and consistently occurs.

81, TITLE:      BOND: BERT-Assisted Open-Domain Named Entity Recognition with Distant Supervision
http://arxiv.org/abs/2006.15509
AUTHORS:        Chen Liang ; Yue Yu ; Haoming Jiang ; Siawpeng Er ; Ruijia Wang ; Tuo Zhao ; Chao Zhang
COMMENTS:       Proceedings of the 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '20)
HIGHLIGHT:      To address this challenge, we propose a new computational framework -- BOND, which leverages the power of pre-trained language models (e.g., BERT and RoBERTa) to improve the prediction performance of NER models.

82, TITLE:      Improving Sequence Tagging for Vietnamese Text Using Transformer-based Neural Models
http://arxiv.org/abs/2006.15994
AUTHORS:        Viet Bui The ; Oanh Tran Thi ; Phuong Le-Hong
HIGHLIGHT:      This paper describes our study on using mutilingual BERT embeddings and some new neural models for improving sequence tagging tasks for the Vietnamese language.

83, TITLE:      Counting Out Time: Class Agnostic Video Repetition Counting in the Wild
http://arxiv.org/abs/2006.15418
AUTHORS:        Debidatta Dwibedi ; Yusuf Aytar ; Jonathan Tompson ; Pierre Sermanet ; Andrew Zisserman
COMMENTS:       Accepted at CVPR 2020. Project webpage: https://sites.google.com/view/repnet
HIGHLIGHT:      We present an approach for estimating the period with which an action is repeated in a video. We also collect a new challenging dataset called Countix (~90 times larger than existing datasets) which captures the challenges of repetition counting in real-world videos.

84, TITLE:      Improving Interpretability of CNN Models Using Non-Negative Concept Activation Vectors
http://arxiv.org/abs/2006.15417
AUTHORS:        Ruihan Zhang ; Prashan Madumal ; Tim Miller ; Kris Ehinger ; Benjamin Rubinstein
HIGHLIGHT:      In this work, we rethink the ACE algorithm of Ghorbani et~al., proposing an alternative concept-based explanation framework.

85, TITLE:      String-based methods for tonal harmony: A corpus study of Haydn's string quartets
http://arxiv.org/abs/2006.15411
AUTHORS:        David R. W. Sears
COMMENTS:       This is an original manuscript / preprint of a book chapter: Sears, David R. W (in press). String-based methods for tonal harmony: A corpus study of Haydn's string quartets." In D. Shanahan, A. Burgoyne, & I. Quinn (Eds.), Oxford Handbook of Music and Corpus Studies. New York: Oxford University Press. The manuscript contains 2 musical examples, 3 figures, and 4 tables
HIGHLIGHT:      This chapter considers how string-based methods might be adapted to address music-analytic questions related to the discovery of musical organization, with particular attention devoted to the analysis of tonal harmony.

86, TITLE:      Enhancement of a CNN-Based Denoiser Based on Spatial and Spectral Analysis
http://arxiv.org/abs/2006.15517
AUTHORS:        Rui Zhao ; Kin-Man Lam ; Daniel P. K. Lun
COMMENTS:       ICIP 2019
HIGHLIGHT:      In this paper, we propose a discrete wavelet denoising CNN (WDnCNN), which restores images corrupted by various noise with a single model.

87, TITLE:      On the generalization of learning-based 3D reconstruction
http://arxiv.org/abs/2006.15427
AUTHORS:        Miguel Angel Bautista ; Walter Talbott ; Shuangfei Zhai ; Nitish Srivastava ; Joshua M Susskind
HIGHLIGHT:      In this paper we study the inductive biases encoded in the model architecture that impact the generalization of learning-based 3D reconstruction methods.

88, TITLE:      Predictive and Generative Neural Networks for Object Functionality
http://arxiv.org/abs/2006.15520
AUTHORS:        Ruizhen Hu ; Zihao Yan ; Jingwen Zhang ; Oliver van Kaick ; Ariel Shamir ; Hao Zhang ; Hui Huang
COMMENTS:       Accepted to SIGGRAPH 2018, project page at https://vcc.tech/research/2018/ICON4
HIGHLIGHT:      We develop predictive and generative deep convolutional neural networks to replicate this feat.

89, TITLE:      DeepACC:Automate Chromosome Classification based on Metaphase Images using Deep Learning Framework Fused with Prior Knowledge
http://arxiv.org/abs/2006.15528
AUTHORS:        Chunlong Luo ; Tianqi Yu ; Yufan Luo ; Manqing Wang ; Fuhai Yu ; Yinhao Li ; Chan Tian ; Jie Qiao ; Li Xiao
HIGHLIGHT:      In this work, we propose a detection based method, DeepACC, to locate and fine classify chromosomes simultaneously based on the whole metaphase image.

90, TITLE:      Few-Shot Class-Incremental Learning via Feature Space Composition
http://arxiv.org/abs/2006.15524
AUTHORS:        Hanbin Zhao ; Yongjian Fu ; Xuewei Li ; Songyuan Li ; Bourahla Omar ; Xi Li
HIGHLIGHT:      With this motivation, we propose a novel few-shot class-incremental learning pipeline based on a composite representation space, which makes old-knowledge preserving and new-knowledge adaptation mutually compatible by feature space composition (enlarging the embedding capacity).

91, TITLE:      Compositional Convolutional Neural Networks: A Robust and Interpretable Model for Object Recognition under Occlusion
http://arxiv.org/abs/2006.15538
AUTHORS:        Adam Kortylewski ; Qing Liu ; Angtian Wang ; Yihong Sun ; Alan Yuille
HIGHLIGHT:      In this work, we show that black-box deep convolutional neural networks (DCNNs) have only limited robustness to partial occlusion.

92, TITLE:      Mind The Facts: Knowledge-Boosted Coherent Abstractive Text Summarization
http://arxiv.org/abs/2006.15435
AUTHORS:        Beliz Gunel ; Chenguang Zhu ; Michael Zeng ; Xuedong Huang
COMMENTS:       NeurIPS 2019, Knowledge Representation & Reasoning Meets Machine Learning (KR2ML workshop)
HIGHLIGHT:      In this work, we propose a novel architecture that extends Transformer encoder-decoder architecture in order to improve on these shortcomings.

93, TITLE:      DHARI Report to EPIC-Kitchens 2020 Object Detection Challenge
http://arxiv.org/abs/2006.15553
AUTHORS:        Kaide Li ; Bingyan Liao ; Laifeng Hu ; Yaonong Wang
COMMENTS:       Challenge Winner in the EPIC-Kitchens 2020 Object Detection Challenge(EPIC@CVPR2020)
HIGHLIGHT:      Experimental results demonstrate that our approach can significantly improve the mean Average Precision (mAP) of object detection on both the seen and unseen test sets of EPICKitchens.

94, TITLE:      SAR Image Despeckling by Deep Neural Networks: from a pre-trained model to an end-to-end training strategy
http://arxiv.org/abs/2006.15559
AUTHORS:        Emanuele Dalsasso ; Loïc Denis ; Florence Tupin
COMMENTS:       Notebook with Colab compatibility is available at https://github.com/emanueledalsasso/SAR-CNN
HIGHLIGHT:      To handle this problem, this paper analyzes different strategies one can adopt, depending on the speckle removal task one wishes to perform and the availability of multitemporal stacks of SAR data.

95, TITLE:      When and How Can Deep Generative Models be Inverted?
http://arxiv.org/abs/2006.15555
AUTHORS:        Aviad Aberdam ; Dror Simon ; Michael Elad
HIGHLIGHT:      Deep generative models (e.g. GANs and VAEs) have been developed quite extensively in recent years.

96, TITLE:      I can attend a meeting too! Towards a human-like telepresence avatar robot to attend meeting on your behalf
http://arxiv.org/abs/2006.15647
AUTHORS:        Hrishav Bakul Barua ; Chayan Sarkar ; Achanna Anil Kumar ; Arpan Pal ; Balamuralidhar P
HIGHLIGHT:      In this work, we focus on a telepresence robot that can be used to attend a meeting remotely with a group of people.

97, TITLE:      Video Representations of Goals Emerge from Watching Failure
http://arxiv.org/abs/2006.15657
AUTHORS:        Dave Epstein ; Carl Vondrick
HIGHLIGHT:      We introduce a video representation learning framework that models the latent goals behind observable human action.

98, TITLE:      Geometry-Inspired Top-k Adversarial Perturbations
http://arxiv.org/abs/2006.15669
AUTHORS:        Nurislam Tursynbek ; Aleksandr Petiushko ; Ivan Oseledets
HIGHLIGHT:      We propose an effective geometry-inspired method of computing Top-k adversarial examples for any k.

99, TITLE:      Differential Privacy of Hierarchical Census Data: An Optimization Approach
http://arxiv.org/abs/2006.15673
AUTHORS:        Ferdinando Fioretto ; Pascal Van Hentenryck ; Keyu Zhu
HIGHLIGHT:      The key technical contribution of the paper shows that this optimization problem can be solved in polynomial time by exploiting the structure of its cost functions.

100, TITLE:     A Retinex based GAN Pipeline to Utilize Paired and Unpaired Datasets for Enhancing Low Light Images
http://arxiv.org/abs/2006.15304
AUTHORS:        Harshana Weligampola ; Gihan Jayatilaka ; Suren Sritharan ; Roshan Godaliyadda ; Parakrama Ekanayaka ; Roshan Ragel ; Vijitha Herath
HIGHLIGHT:      This paper presents a novel deep learning pipeline that can learn from both paired and unpaired datasets.

101, TITLE:     A Benchmark dataset for both underwater image enhancement and underwater object detection
http://arxiv.org/abs/2006.15789
AUTHORS:        Long Chen ; Lei Tong ; Feixiang Zhou ; Zheheng Jiang ; Zhenyang Li ; Jialin Lv ; Junyu Dong ; Huiyu Zhou
HIGHLIGHT:      To investigate how the underwater image enhancement methods influence the following underwater object detection tasks, in this paper, we provide a large-scale underwater object detection dataset with both bounding box annotations and high quality reference images, namely OUC dataset.

102, TITLE:     Seasonal Averaged One-Dependence Estimators: A Novel Algorithm to Address Seasonal Concept Drift in High-Dimensional Stream Classification
http://arxiv.org/abs/2006.15311
AUTHORS:        Rakshitha Godahewa ; Trevor Yann ; Christoph Bergmeir ; Francois Petitjean
HIGHLIGHT:      We introduce a novel classifier named Seasonal Averaged One-Dependence Estimators (SAODE), which extends the AODE classifier to handle seasonal drift by including time as a super parent.

103, TITLE:     Combine Convolution with Recurrent Networks for Text Classification
http://arxiv.org/abs/2006.15795
AUTHORS:        Shengfei Lyu ; Jiaqi Liu
HIGHLIGHT:      In this paper, we propose a novel method to keep the strengths of the two networks to a great extent.

104, TITLE:     Uncertainty-aware Self-training for Text Classification with Few Labels
http://arxiv.org/abs/2006.15315
AUTHORS:        Subhabrata Mukherjee ; Ahmed Hassan Awadallah
HIGHLIGHT:      In this work, we study self-training as one of the earliest semi-supervised learning approaches to reduce the annotation bottleneck by making use of large-scale unlabeled data for the target task.

105, TITLE:     Video-Grounded Dialogues with Pretrained Generation Language Models
http://arxiv.org/abs/2006.15319
AUTHORS:        Hung Le ; Steven C. H. Hoi
COMMENTS:       Accepted at ACL 2020 (Short Paper)
HIGHLIGHT:      In this paper, we leverage the power of pre-trained language models for improving video-grounded dialogue, which is very challenging and involves complex features of different dynamics: (1) Video features which can extend across both spatial and temporal dimensions; and (2) Dialogue features which involve semantic dependencies over multiple dialogue turns.

106, TITLE:     Interactive Deep Refinement Network for Medical Image Segmentation
http://arxiv.org/abs/2006.15320
AUTHORS:        Titinunt Kitrungrotsakul ; Iwamoto Yutaro ; Lanfen Lin ; Ruofeng Tong ; Jingsong Li ; Yen-Wei Chen
COMMENTS:       10 pages, 4 figures
HIGHLIGHT:      In this paper, we propose an interactive deep refinement framework to improve the traditional semantic segmentation networks such as U-Net and fully convolutional network.

107, TITLE:     Dynamic Sampling Networks for Efficient Action Recognition in Videos
http://arxiv.org/abs/2006.15560
AUTHORS:        Yin-Dong Zheng ; Zhaoyang Liu ; Tong Lu ; Limin Wang
COMMENTS:       To appear in IEEE Transaction on Image Processing
HIGHLIGHT:      To address these issues, we propose a new framework for action recognition in videos, called {\em Dynamic Sampling Networks} (DSN), by designing a dynamic sampling module to improve the discriminative power of learned clip-level classifiers and as well increase the inference efficiency during testing.

108, TITLE:     Joint Hand-object 3D Reconstruction from a Single Image with Cross-branch Feature Fusion
http://arxiv.org/abs/2006.15561
AUTHORS:        Yujin Chen ; Zhigang Tu ; Di Kang ; Ruizhi Chen ; Linchao Bao ; Zhengyou Zhang ; Junsong Yuan
HIGHLIGHT:      In this work, we propose to consider hand and object jointly in feature space and explore the reciprocity of the two branches.

109, TITLE:     Data augmentation versus noise compensation for x- vector speaker recognition systems in noisy environments
http://arxiv.org/abs/2006.15903
AUTHORS:        Mohammad Mohammadamini ; Driss Matrouf
HIGHLIGHT:      In this work, we want to know if explicit noise compensation techniques continue to be effective despite the general noise robustness of these systems.

110, TITLE:     Using Reinforcement Learning to Herd a Robotic Swarm to a Target Distribution
http://arxiv.org/abs/2006.15807
AUTHORS:        Zahi M. Kakish ; Karthik Elamvazhuthi ; Spring Berman
COMMENTS:       Paper was submitted to Conference on Robot Learning 2019 and IEEE Robotics and Automation Letters 2020
HIGHLIGHT:      In this paper, we present a reinforcement learning approach to designing a control policy for a "leader'' agent that herds a swarm of "follower'' agents, via repulsive interactions, as quickly as possible to a target probability distribution over a strongly connected graph.

111, TITLE:     Revision by Conditionals: From Hook to Arrow
http://arxiv.org/abs/2006.15811
AUTHORS:        Jake Chandler ; Richard Booth
COMMENTS:       Extended version of a paper accepted to KR 2020
HIGHLIGHT:      We introduce a 'plug and play' method for uniquely extending any iterated belief revision operator to the conditional case.

112, TITLE:     Improving Few-Shot Learning using Composite Rotation based Auxiliary Task
http://arxiv.org/abs/2006.15919
AUTHORS:        Pratik Mazumder ; Pravendra Singh ; Vinay P. Namboodiri
HIGHLIGHT:      In this paper, we propose an approach to improve few-shot classification performance using a composite rotation based auxiliary task.

113, TITLE:     Interpreting and Disentangling Feature Components of Various Complexity from DNNs
http://arxiv.org/abs/2006.15920
AUTHORS:        Jie Ren ; Mingjie Li ; Zexu Liu ; Quanshi Zhang
HIGHLIGHT:      This paper aims to define, quantify, and analyze the feature complexity that is learned by a DNN.

114, TITLE:     Retro*: Learning Retrosynthetic Planning with Neural Guided A* Search
http://arxiv.org/abs/2006.15820
AUTHORS:        Binghong Chen ; Chengtao Li ; Hanjun Dai ; Le Song
COMMENTS:       Presented at ICML 2020
HIGHLIGHT:      In this paper, we propose Retro*, a neural-based A*-like algorithm that finds high-quality synthetic routes efficiently.

115, TITLE:     Towards Learning-automation IoT Attack Detection through Reinforcement Learning
http://arxiv.org/abs/2006.15826
AUTHORS:        Tianbo Gu ; Allaukik Abhishek ; Hao Fu ; Huanle Zhang ; Debraj Basu ; Prasant Mohapatra
COMMENTS:       11 pages, 8 figures, 2 tables, to appear in the 21st IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks (IEEE WoWMoM 2020)
HIGHLIGHT:      In order to adapt to the new characteristics in IoT attacks, we propose a reinforcement learning-based attack detection model that can automatically learn and recognize the transformation of the attack pattern.

116, TITLE:     Solving MKP Applied to IoT in Smart Grid Using Meta-heuristics Algorithms: A Parallel Processing Perspective
http://arxiv.org/abs/2006.15927
AUTHORS:        Jandre Albertyn ; Ling Cheng ; Adnan M. Abu-Mahfouz
HIGHLIGHT:      Solving MKP Applied to IoT in Smart Grid Using Meta-heuristics Algorithms: A Parallel Processing Perspective

117, TITLE:     Answering Questions on COVID-19 in Real-Time
http://arxiv.org/abs/2006.15830
AUTHORS:        Jinhyuk Lee ; Sean S. Yi ; Minbyul Jeong ; Mujeen Sung ; Wonjin Yoon ; Yonghwa Choi ; Miyoung Ko ; Jaewoo Kang
COMMENTS:       10 pages
HIGHLIGHT:      In this work, we outline our effort to contribute to shrinking this knowledge vacuum by creating covidAsk, a question answering (QA) system that combines biomedical text mining and QA techniques to provide answers to questions in real-time.

118, TITLE:     Is Japanese gendered language used on Twitter ? A large scale study
http://arxiv.org/abs/2006.15935
AUTHORS:        Tiziana Carpi ; Stefano Maria Iacus
HIGHLIGHT:      This study analyzes the usage of Japanese gendered language on Twitter.

119, TITLE:     Hybrid Tensor Decomposition in Neural Network Compression
http://arxiv.org/abs/2006.15938
AUTHORS:        Bijiao Wu ; Dingheng Wang ; Guangshe Zhao ; Lei Deng ; Guoqi Li
COMMENTS:       submitted to <<Neural Networks>> on Apr.18,2020; received first review comments on June.02,2020; the revised manuscipt submitted on June.28,2020
HIGHLIGHT:      In this work, we introduce the hierarchical Tucker (HT), a classical but rarely-used tensor decomposition method, to investigate its capability in neural network compression.

120, TITLE:     End-to-End Differentiable Learning to HDR Image Synthesis for Multi-exposure Images
http://arxiv.org/abs/2006.15833
AUTHORS:        Jung Hee Kim ; Siyeong Lee ; Soyeon Jo ; Suk-Ju Kang
COMMENTS:       10 pages, 5 figures
HIGHLIGHT:      Therefore, we tackle the major challenge in stack reconstruction-based methods by proposing a novel framework with the fully differentiable HDRI process.

121, TITLE:     Adversarial Multi-Source Transfer Learning in Healthcare: Application to Glucose Prediction for Diabetic People
http://arxiv.org/abs/2006.15940
AUTHORS:        Maxime De Bois ; Mounîm A. El Yacoubi ; Mehdi Ammi
HIGHLIGHT:      To improve the quality of the transfer between multiple sources of data, we propose a multi-source adversarial transfer learning framework that enables the learning of a feature representation that is similar across the sources, and thus more general and more easily transferable.

122, TITLE:     Hinting Semantic Parsing with Statistical Word Sense Disambiguation
http://arxiv.org/abs/2006.15942
AUTHORS:        Ritwik Bose ; Siddharth Vashishtha ; James Allen
COMMENTS:       7 pages, 3 figures
HIGHLIGHT:      In this work, we provide hints from a statistical WSD system to guide a logical semantic parser to produce better semantic type assignments while maintaining the soundness of the resulting logical forms.

123, TITLE:     Towards hybrid primary intersubjectivity: a neural robotics library for human science
http://arxiv.org/abs/2006.15948
AUTHORS:        Hendry F. Chame ; Ahmadreza Ahmadi ; Jun Tani
HIGHLIGHT:      Hence, in this work we pursue three main objectives.

124, TITLE:     A Transformer-based joint-encoding for Emotion Recognition and Sentiment Analysis
http://arxiv.org/abs/2006.15955
AUTHORS:        Jean-Benoit Delbrouck ; Noé Tits ; Mathilde Brousmiche ; Stéphane Dupont
COMMENTS:       Winner of the ACL20: Second Grand-Challenge on Multimodal Language
HIGHLIGHT:      This paper describes a Transformer-based joint-encoding (TBJE) for the task of Emotion Recognition and Sentiment Analysis.

125, TITLE:     Multi-level colonoscopy malignant tissue detection with adversarial CAC-UNet
http://arxiv.org/abs/2006.15954
AUTHORS:        Chuang Zhu ; Ke Mei ; Ting Peng ; Yihao Luo ; Jun Liu ; Ying Wang ; Mulan Jin
COMMENTS:       accepted by Neurocomputing; winner of the MICCAI DigestPath 2019 challenge on colonoscopy tissue segmentation and classification task
HIGHLIGHT:      In this paper, we propose a highly efficient multi-level malignant tissue detection through the designed adversarial CAC-UNet.

126, TITLE:     Application of Neuroevolution in Autonomous Cars
http://arxiv.org/abs/2006.15175
AUTHORS:        Sainath G ; Vignesh S ; Siddarth S ; G Suganya
COMMENTS:       13 pages, 9 figures, 1 table
HIGHLIGHT:      We propose a system that requires no data for its training.

127, TITLE:     Bookworm continual learning: beyond zero-shot learning and continual learning
http://arxiv.org/abs/2006.15176
AUTHORS:        Kai Wang ; Luis Herranz ; Anjan Dutta ; Joost van de Weijer
HIGHLIGHT:      We propose bookworm continual learning(BCL), a flexible setting where unseen classes can be inferred via a semantic model, and the visual model can be updated continually.

128, TITLE:     Creating Artificial Modalities to Solve RGB Liveness
http://arxiv.org/abs/2006.16028
AUTHORS:        Aleksandr Parkin ; Oleg Grinchuk
COMMENTS:       CVPRW2020
HIGHLIGHT:      In this work we propose a method to utilize the difference in dynamic appearance between bona fide and spoof samples by creating artificial modalities from RGB videos.

129, TITLE:     Concept and the implementation of a tool to convert industry 4.0 environments modeled as FSM to an OpenAI Gym wrapper
http://arxiv.org/abs/2006.16035
AUTHORS:        Kallil M. C. Zielinski ; Marcelo Teixeira ; Richardson Ribeiro ; Dalcimar Casanova
HIGHLIGHT:      In this way, this work presents the concept and the implementation of a tool that allows us to convert any dynamic system modeled as an FSM to the open-source Gym wrapper.

130, TITLE:     Visual Kinship Recognition: A Decade in the Making
http://arxiv.org/abs/2006.16033
AUTHORS:        Joseph P Robinson ; Ming Shao ; Yun Fu
HIGHLIGHT:      We list and review the public resources and data challenges that enabled and inspired many to hone-in on one or more views of automatic kinship recognition in the visual domain.

131, TITLE:     Region-of-interest guided Supervoxel Inpainting for Self-supervision
http://arxiv.org/abs/2006.15186
AUTHORS:        Subhradeep Kayal ; Shuai Chen ; Marleen de Bruijne
COMMENTS:       Accepted at MICCAI 2020
HIGHLIGHT:      In this work, we focus on image inpainting as the self-supervised proxy task, and propose two novel structural changes to further enhance the performance of a deep neural network.

132, TITLE:     Making DensePose fast and light
http://arxiv.org/abs/2006.15190
AUTHORS:        Ruslan Rakhimov ; Emil Bogomolov ; Alexandr Notchenko ; Fung Mao ; Alexey Artemov ; Denis Zorin ; Evgeny Burnaev
HIGHLIGHT:      In this work, we target the problem of redesigning the DensePose R-CNN model's architecture so that the final network retains most of its accuracy but becomes more light-weight and fast.

133, TITLE:     Automatic Operating Room Surgical Activity Recognition for Robot-Assisted Surgery
http://arxiv.org/abs/2006.16166
AUTHORS:        Aidean Sharghi ; Helene Haugerud ; Daniel Oh ; Omid Mohareri
COMMENTS:       International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI'20)
HIGHLIGHT:      In this paper, we investigate automatic surgical activity recognition in robot-assisted operations. We collect the first large-scale dataset including 400 full-length multi-perspective videos from a variety of robotic surgery cases captured using Time-of-Flight cameras.

134, TITLE:     Building Rule Hierarchies for Efficient Logical Rule Learning from Knowledge Graphs
http://arxiv.org/abs/2006.16171
AUTHORS:        Yulong Gu ; Yu Guan ; Paolo Missior
HIGHLIGHT:      In this work, we address such scalability issues by proposing new methods for pruning unpromising rules using rule hierarchies.

135, TITLE:     Multichannel CNN with Attention for Text Classification
http://arxiv.org/abs/2006.16174
AUTHORS:        Zhenyu Liu ; Haiwei Huang ; Chaohong Lu ; Shengfei Lyu
HIGHLIGHT:      In order to combine the strengths of the two kinds of networks and alleviate their shortcomings, this paper proposes Attention-based Multichannel Convolutional Neural Network (AMCNN) for text classification.

136, TITLE:     Natural Backdoor Attack on Text Data
http://arxiv.org/abs/2006.16176
AUTHORS:        Lichao Sun
COMMENTS:       Paper submitted to a conference, due to the double-blind policy, we only preprint it with all random name and school information
HIGHLIGHT:      In this paper, we systematically study the backdoor attack against models on text data.

137, TITLE:     Unsupervised Learning Consensus Model for Dynamic Texture Videos Segmentation
http://arxiv.org/abs/2006.16177
AUTHORS:        Lazhar Khelifi ; Max Mignotte
HIGHLIGHT:      We present an effective unsupervised learning consensus model for the segmentation of dynamic texture (ULCM).

138, TITLE:     Harvesting, Detecting, and Characterizing Liver Lesions from Large-scale Multi-phase CT Data via Deep Dynamic Texture Learning
http://arxiv.org/abs/2006.15691
AUTHORS:        Yuankai Huo ; Jinzheng Cai ; Chi-Tung Cheng ; Ashwin Raju ; Ke Yan ; Bennett A. Landman ; Jing Xiao ; Le Lu ; Chien-Hung Liao ; Adam Harrison
HIGHLIGHT:      In this work, we curate a patient cohort of 1305 dynamic contrast CT studies (i.e., 5220 multi-phase 3D volumes) with pathology confirmed ground truth.

139, TITLE:     Simulation of Brain Resection for Cavity Segmentation Using Self-Supervised and Semi-Supervised Learning
http://arxiv.org/abs/2006.15693
AUTHORS:        Fernando Pérez-García ; Roman Rodionov ; Ali Alim-Marvasti ; Rachel Sparks ; John S. Duncan ; Sébastien Ourselin
COMMENTS:       13 pages, 6 figures, accepted at the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2020
HIGHLIGHT:      We developed an algorithm to simulate resections on preoperative MR images.

140, TITLE:     Traditional and accelerated gradient descent for neural architecture search
http://arxiv.org/abs/2006.15218
AUTHORS:        Nicolas Garcia Trillos ; Felix Morales ; Javier Morales
HIGHLIGHT:      In this paper, we introduce two algorithms for neural architecture search (NASGD and NASAGD) following the theoretical work by two of the authors [3].

141, TITLE:     MIMC-VINS: A Versatile and Resilient Multi-IMU Multi-Camera Visual-Inertial Navigation System
http://arxiv.org/abs/2006.15699
AUTHORS:        Kevin Eckenhoff ; Patrick Geneva ; Guoquan Huang
COMMENTS:       20 pages, 10 figures, 13 tables
HIGHLIGHT:      To this end, rather than the standard VINS paradigm using a minimal sensing suite of a single camera and IMU, in this paper we design a real-time consistent multi-IMU multi-camera (MIMC)-VINS estimator that is able to seamlessly fuse multi-modal information from an arbitrary number of uncalibrated cameras and IMUs.



==========Updates to Previous Papers==========
1, TITLE:       Channel Pruning via Automatic Structure Search
http://arxiv.org/abs/2001.08565
AUTHORS:        Mingbao Lin ; Rongrong Ji ; Yuxin Zhang ; Baochang Zhang ; Yongjian Wu ; Yonghong Tian
COMMENTS:       Accepted by IJCAI2020. SOLO copyright holder is IJCAI (International Joint Conferences on Artificial Intelligence)
HIGHLIGHT:      In this paper, we propose a new channel pruning method based on artificial bee colony algorithm (ABC), dubbed as ABCPruner, which aims to efficiently find optimal pruned structure, i.e., channel number in each layer, rather than selecting "important" channels as previous works did.

2, TITLE:       Cross-Supervised Object Detection
http://arxiv.org/abs/2006.15056
AUTHORS:        Zitian Chen ; Zhiqiang Shen ; Jiahui Yu ; Erik Learned-Miller
HIGHLIGHT:      In this work, we show how to build better object detectors from weakly labeled images of new categories by leveraging knowledge learned from fully labeled base categories.

3, TITLE:       A Novel Nudity Detection Algorithm for Web and Mobile Application Development
http://arxiv.org/abs/2006.01780
AUTHORS:        Rahat Yeasin Emon
COMMENTS:       5 pages
HIGHLIGHT:      This paper presents a runtime nudity detection method for web and mobile application development.

4, TITLE:       FastReID: A Pytorch Toolbox for General Instance Re-identification
http://arxiv.org/abs/2006.02631
AUTHORS:        Lingxiao He ; Xingyu Liao ; Wu Liu ; Xinchen Liu ; Peng Cheng ; Tao Mei
HIGHLIGHT:      To meet the increasing application demand for general instance re-identification, we present FastReID as a widely used software system in JD AI Research.

5, TITLE:       POMDP Modelling for Assessing Hierarchies
http://arxiv.org/abs/1908.07031
AUTHORS:        Weipeng Huang ; Guangyuan Piao ; Raul Moreno ; Neil J. Hurley
HIGHLIGHT:      This motivates us to propose a framework for assessing the quality of hierarchical clustering allocations which covers the case of no ground-truth information.

6, TITLE:       Exploring the Capacity of an Orderless Box Discretization Network for Multi-orientation Scene Text Detection
http://arxiv.org/abs/1912.09629
AUTHORS:        Yuliang Liu ; Tong He ; Hao Chen ; Xinyu Wang ; Canjie Luo ; Shuaitao Zhang ; Chunhua Shen ; Lianwen Jin
HIGHLIGHT:      Here we solve this problem by proposing a novel method, termed Orderless Box Discretization (OBD), which first discretizes the quadrilateral box into several key edges containing all potential horizontal and vertical positions.

7, TITLE:       Universal Self-Attention Network for Graph Classification
http://arxiv.org/abs/1909.11855
AUTHORS:        Dai Quoc Nguyen ; Tu Dinh Nguyen ; Dinh Phung
COMMENTS:       We have updated the Pytorch and Tensorflow implementation at: https://github.com/daiquocnguyen/Graph-Transformer
HIGHLIGHT:      To this end, we present U2GNN -- a novel embedding model leveraging the transformer self-attention network -- to learn plausible node and graph embeddings.

8, TITLE:       Efficient Second-Order TreeCRF for Neural Dependency Parsing
http://arxiv.org/abs/2005.00975
AUTHORS:        Yu Zhang ; Zhenghua Li ; Min Zhang
COMMENTS:       ACL 2020
HIGHLIGHT:      To address this issue, we propose an effective way to batchify the inside and Viterbi algorithms for direct large matrix operation on GPUs, and to avoid the complex outside algorithm via efficient back-propagation.

9, TITLE:       Symbolic Querying of Vector Spaces: Probabilistic Databases Meets Relational Embeddings
http://arxiv.org/abs/2002.10029
AUTHORS:        Tal Friedman ; Guy Van den Broeck
HIGHLIGHT:      We propose unifying techniques from probabilistic databases and relational embedding models with the goal of performing complex queries on incomplete and uncertain data.

10, TITLE:      "Good Robot!": Efficient Reinforcement Learning for Multi-Step Visual Tasks with Sim to Real Transfer
http://arxiv.org/abs/1909.11730
AUTHORS:        Andrew Hundt ; Benjamin Killeen ; Nicholas Greene ; Hongtao Wu ; Heeyeon Kwon ; Chris Paxton ; Gregory D. Hager
COMMENTS:       This is a major update to v2 with 100% success in real robot tests, clearer equations, a comprehensive ablation study, and generalization to a grid world environment. 8 pages, 6 figures, 3 tables, 1 algorithm. Code is available at https://github.com/jhu-lcsr/good_robot and a video overview is at https://youtu.be/MbCuEZadkIw
HIGHLIGHT:      The SPOT framework successfully completes simulated trials of a variety of tasks, improving a baseline trial success rate from 13% to 100% when stacking 4 cubes, from 13% to 99% when creating rows of 4 cubes, and from 84% to 95% when clearing toys arranged in adversarial patterns.

11, TITLE:      It's Not What Machines Can Learn, It's What We Cannot Teach
http://arxiv.org/abs/2002.09398
AUTHORS:        Gal Yehuda ; Moshe Gabel ; Assaf Schuster
COMMENTS:       Accepted to ICML 2020
HIGHLIGHT:      In this work we offer a different perspective on this question.

12, TITLE:      Joint Spatial and Angular Super-Resolution from a Single Image
http://arxiv.org/abs/1911.11619
AUTHORS:        Andre Ivan ;  Williem ; In Kyu Park
COMMENTS:       arXiv admin note: substantial text overlap with arXiv:1903.12364
HIGHLIGHT:      In this paper, we show that both super-resolution problems can be solved jointly from a single image by proposing a single end-to-end deep neural network that does not require a physical-based approach.

13, TITLE:      2D Image Relighting with Image-to-Image Translation
http://arxiv.org/abs/2006.07816
AUTHORS:        Paul Gafton ; Erick Maraz
COMMENTS:       12 pages, 52 Postscript figures, uses cvpr_eso.sty eso-pic.sty ruler.sty
HIGHLIGHT:      Here we provide our attempt to solve this problem using GANs.

14, TITLE:      A Novel Approach for Correcting Multiple Discrete Rigid In-Plane Motions Artefacts in MRI Scans
http://arxiv.org/abs/2006.13804
AUTHORS:        Michael Rotman ; Rafi Brada ; Israel Beniaminy ; Sangtae Ahn ; Christopher J. Hardy ; Lior Wolf
HIGHLIGHT:      In this paper we propose a novel method for removing motion artefacts using a deep neural network with two input branches that discriminates between patient poses using the motion's timing.

15, TITLE:      G2D: Generate to Detect Anomaly
http://arxiv.org/abs/2006.11629
AUTHORS:        Masoud Pourreza ; Bahram Mohammadi ; Mostafa Khaki ; Samir Bouindour ; Hichem Snoussi ; Mohammad Sabokrou
HIGHLIGHT:      In this paper, we propose a novel method for irregularity detection.

16, TITLE:      Failure of Normalization in Impredicative Type Theory with Proof-Irrelevant Propositional Equality
http://arxiv.org/abs/1911.08174
AUTHORS:        Andreas Abel ; Thierry Coquand
HIGHLIGHT:      Failure of Normalization in Impredicative Type Theory with Proof-Irrelevant Propositional Equality

17, TITLE:      On the Relationship Between Active Inference and Control as Inference
http://arxiv.org/abs/2006.12964
AUTHORS:        Beren Millidge ; Alexander Tschantz ; Anil K Seth ; Christopher L Buckley
COMMENTS:       final workshop version
HIGHLIGHT:      In the context of this comparison, we highlight several ways in which these frameworks can inform one another.

18, TITLE:      Context-Sensitive Generation Network for Handing Unknown Slot Values in Dialogue State Tracking
http://arxiv.org/abs/2005.03923
AUTHORS:        Puhai Yang ; Heyan Huang ; Xian-Ling Mao
HIGHLIGHT:      To tackle the problem, in this paper, we propose a novel Context-Sensitive Generation network (CSG) which can facilitate the representation of out-of-vocabulary words when generating the unknown slot value.

19, TITLE:      Self-supervised Robust Object Detectors from Partially Labelled Datasets
http://arxiv.org/abs/2005.11549
AUTHORS:        Mahdieh Abbasi ; Denis Laurendeau ; Christian Gagne
HIGHLIGHT:      With the goal of training \emph{one integrated robust object detector with high generalization performance}, we propose a training framework to overcome missing-label challenge of the merged datasets.

20, TITLE:      Dynamic Task Weighting Methods for Multi-task Networks in Autonomous Driving Systems
http://arxiv.org/abs/2001.02223
AUTHORS:        Isabelle Leang ; Ganesh Sistu ; Fabian Burger ; Andrei Bursuc ; Senthil Yogamani
COMMENTS:       Accepted for Oral Presentation at IEEE Intelligent Transportation Systems Conference (ITSC) 2020
HIGHLIGHT:      In this work, we review and systematically evaluate nine task weighting strategies on common grounds on three automotive datasets (KITTI, Cityscapes and WoodScape).

21, TITLE:      Two-phase protein folding optimization on a three-dimensional AB off-lattice model
http://arxiv.org/abs/1903.01456
AUTHORS:        Borko Bošković ; Janez Brest
COMMENTS:       20 pages, 13 tables, 6 figures. arXiv admin note: text overlap with arXiv:1710.07031
HIGHLIGHT:      This paper presents a two-phase protein folding optimization on a three-dimensional AB off-lattice model.

22, TITLE:      A novel approach for multi-agent cooperative pursuit to capture grouped evaders
http://arxiv.org/abs/2006.01022
AUTHORS:        Muhammad Zuhair Qadir ; Songhao Piao ; Haiyang Jiang ; Mohammed El Habib Souidi
COMMENTS:       published paper's draft version
HIGHLIGHT:      An approach of mobile multi-agent pursuit based on application of self-organizing feature map (SOFM) and along with that reinforcement learning based on agent group role membership function (AGRMF) model is proposed.

23, TITLE:      Transformer based Grapheme-to-Phoneme Conversion
http://arxiv.org/abs/2004.06338
AUTHORS:        Sevinj Yolchuyeva ; Géza Németh ; Bálint Gyires-Tóth
COMMENTS:       INTERSPEECH 2019
HIGHLIGHT:      In this paper, we investigate the application of transformer architecture to G2P conversion and compare its performance with recurrent and convolutional neural network based approaches.

24, TITLE:      Adaptive Online Planning for Continual Lifelong Learning
http://arxiv.org/abs/1912.01188
AUTHORS:        Kevin Lu ; Igor Mordatch ; Pieter Abbeel
COMMENTS:       Originally published in NeurIPS Deep RL 2019
HIGHLIGHT:      We present a new algorithm, Adaptive Online Planning (AOP), that achieves strong performance in this setting by combining model-based planning with model-free learning.

25, TITLE:      Natural Language Processing Advancements By Deep Learning: A Survey
http://arxiv.org/abs/2003.01200
AUTHORS:        Amirsina Torfi ; Rouzbeh A. Shirvani ; Yaser Keneshloo ; Nader Tavaf ; Edward A. Fox
HIGHLIGHT:      It covers core NLP tasks and applications and describes how deep learning methods and models advance these areas.

26, TITLE:      Rapid trial-and-error learning with simulation supports flexible tool use and physical reasoning
http://arxiv.org/abs/1907.09620
AUTHORS:        Kelsey R. Allen ; Kevin A. Smith ; Joshua B. Tenenbaum
COMMENTS:       This manuscript is in press at PNAS. It is an extended version of a paper "Rapid Trial-and-Error Learning in Physical Problem Solving" accepted for oral presentation at the 41st Annual Meeting of the Cognitive Science Society (2019). It represents ongoing work on the part of the authors
HIGHLIGHT:      To study this type of general physical problem solving, we introduce the Virtual Tools game.

27, TITLE:      Symbolic Network: Generalized Neural Policies for Relational MDPs
http://arxiv.org/abs/2002.07375
AUTHORS:        Sankalp Garg ; Aniket Bajpai ;  Mausam
COMMENTS:       In Proceeding of ICML 2020. Code can be found at https://github.com/dair-iitd/symnet
HIGHLIGHT:      We present SymNet, the first neural approach for solving RMDPs that are expressed in the probabilistic planning language of RDDL.

28, TITLE:      Specular- and Diffuse-reflection-based Face Spoofing Detection for Mobile Devices
http://arxiv.org/abs/1907.12400
AUTHORS:        Akinori F. Ebihara ; Kazuyuki Sakurai ; Hitoshi Imaoka
COMMENTS:       Accepted to IJCB2020
HIGHLIGHT:      Here, we propose an efficient face presentation attack detection (PAD) algorithm that requires minimal hardware and only a small database, making it suitable for resource-constrained devices such as mobile phones.

29, TITLE:      Similarity Learning Networks for Animal Individual Re-Identification -- Beyond the Capabilities of a Human Observer
http://arxiv.org/abs/1902.09324
AUTHORS:        Stefan Schneider ; Graham W. Taylor ; Stefan Linquist ; Stefan C. Kremer
COMMENTS:       9 pages, 4 figures, 3 table. WACV 2020 - Deep Learning for Animal Re-ID Workshop
HIGHLIGHT:      The ability for researchers to re-identify an animal individual upon re-encounter is fundamental for addressing a broad range of questions in the study of population dynamics and community/behavioural ecology.

30, TITLE:      Learning to Detect 3D Objects from Point Clouds in Real Time
http://arxiv.org/abs/2006.01250
AUTHORS:        Abhinav Sagar
COMMENTS:       11 pages
HIGHLIGHT:      In this work, we address the problem of 3D object detection from point cloud data in real time.

31, TITLE:      AraBERT: Transformer-based Model for Arabic Language Understanding
http://arxiv.org/abs/2003.00104
AUTHORS:        Wissam Antoun ; Fady Baly ; Hazem Hajj
COMMENTS:       Proceedings of the Twelfth International Conference on Language Resources and Evaluation (LREC 2020), Marseille, France (2020)
HIGHLIGHT:      In this paper, we pre-trained BERT specifically for the Arabic language in the pursuit of achieving the same success that BERT did for the English language.

32, TITLE:      PnPNet: End-to-End Perception and Prediction with Tracking in the Loop
http://arxiv.org/abs/2005.14711
AUTHORS:        Ming Liang ; Bin Yang ; Wenyuan Zeng ; Yun Chen ; Rui Hu ; Sergio Casas ; Raquel Urtasun
COMMENTS:       CVPR2020
HIGHLIGHT:      Towards this goal we propose PnPNet, an end-to-end model that takes as input sequential sensor data, and outputs at each time step object tracks and their future trajectories.

33, TITLE:      Motion Planning Networks: Bridging the Gap Between Learning-based and Classical Motion Planners
http://arxiv.org/abs/1907.06013
AUTHORS:        Ahmed H. Qureshi ; Yinglong Miao ; Anthony Simeonov ; Michael C. Yip
COMMENTS:       Supplementary material including implementation parameters and project videos are available at https://sites.google.com/view/mpnet/home. This work has been accepted for publication at IEEE Transactions on Robotics
HIGHLIGHT:      To train the MPNet models, we present an active continual learning approach that enables MPNet to learn from streaming data and actively ask for expert demonstrations when needed, drastically reducing data for training.

34, TITLE:      Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learning
http://arxiv.org/abs/1908.02983
AUTHORS:        Eric Arazo ; Diego Ortego ; Paul Albert ; Noel E. O'Connor ; Kevin McGuinness
HIGHLIGHT:      We, conversely, propose to learn from unlabeled data by generating soft pseudo-labels using the network predictions.

35, TITLE:      A Closer Look at Invalid Action Masking in Policy Gradient Algorithms
http://arxiv.org/abs/2006.14171
AUTHORS:        Shengyi Huang ; Santiago Ontañón
COMMENTS:       Preprint. Corrected a major issue of the withdrawn version submitted to NeurIPS 2020
HIGHLIGHT:      In this paper, we show that the standard working mechanism of invalid action masking corresponds to valid policy gradient updates.

36, TITLE:      Derivative Manipulation for General Example Weighting
http://arxiv.org/abs/1905.11233
AUTHORS:        Xinshao Wang ; Elyor Kodirov ; Yang Hua ; Neil M. Robertson
COMMENTS:       Is it the right time to make a change from the design of loss function to the design of derivative directly?
HIGHLIGHT:      Therefore, we propose derivative manipulation (DM), a novel and general example weighting approach for training robust deep models under these adverse conditions.

37, TITLE:      Reliable Fidelity and Diversity Metrics for Generative Models
http://arxiv.org/abs/2002.09797
AUTHORS:        Muhammad Ferjad Naeem ; Seong Joon Oh ; Youngjung Uh ; Yunjey Choi ; Jaejun Yoo
COMMENTS:       First two authors have contributed equally; ICML 2020 accepted
HIGHLIGHT:      In this paper, we show that even the latest version of the precision and recall metrics are not reliable yet.

38, TITLE:      DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation
http://arxiv.org/abs/2006.04868
AUTHORS:        Debesh Jha ; Michael A. Riegler ; Dag Johansen ; Pål Halvorsen ; Håvard D. Johansen
HIGHLIGHT:      To improve the performance of U-Net on various segmentation tasks, we propose a novel architecture called DoubleU-Net, which is a combination of two U-Net architectures stacked on top of each other.

39, TITLE:      Neural Machine Translation for Multilingual Grapheme-to-Phoneme Conversion
http://arxiv.org/abs/2006.14194
AUTHORS:        Alex Sokolov ; Tracy Rohlin ; Ariya Rastrow
COMMENTS:       Published in INTERSPEECH (2019)
HIGHLIGHT:      As an alternative, we present a single end-to-end trained neural G2P model that shares same encoder and decoder across multiple languages.

40, TITLE:      Learning Near Optimal Policies with Low Inherent Bellman Error
http://arxiv.org/abs/2003.00153
AUTHORS:        Andrea Zanette ; Alessandro Lazaric ; Mykel Kochenderfer ; Emma Brunskill
COMMENTS:       Bug fixes in appendix; appears in ICML 2020
HIGHLIGHT:      We study the exploration problem with approximate linear action-value functions in episodic reinforcement learning under the notion of low inherent Bellman error, a condition normally employed to show convergence of approximate value iteration.

41, TITLE:      Reducibility and Statistical-Computational Gaps from Secret Leakage
http://arxiv.org/abs/2005.08099
AUTHORS:        Matthew Brennan ; Guy Bresler
COMMENTS:       175 pages; subsumes preliminary draft arXiv:1908.06130; accepted for presentation at the Conference on Learning Theory (COLT) 2020
HIGHLIGHT:      The insight in this work is that a slight generalization of the planted clique conjecture -- secret leakage planted clique -- gives rise to a variety of new average-case reduction techniques, yielding a web of reductions among problems with very different structure.

42, TITLE:      Correlation decay and partition function zeros: Algorithms and phase transitions
http://arxiv.org/abs/1906.01228
AUTHORS:        Jingcheng Liu ; Alistair Sinclair ; Piyush Srivastava
COMMENTS:       New results and discussion added
HIGHLIGHT:      We explore connections between the phenomenon of correlation decay and the location of Lee-Yang and Fisher zeros for various spin systems.

43, TITLE:      Channel Pruning via Optimal Thresholding
http://arxiv.org/abs/2003.04566
AUTHORS:        Yun Ye ; Ganmei You ; Jong-Kae Fwu ; Xia Zhu ; Qing Yang ; Yuan Zhu
COMMENTS:       9 pages, 9 figures, 4 tables
HIGHLIGHT:      In this paper, we present a simple yet effective method, termed Optimal Thresholding (OT), to prune channels with layer dependent thresholds that optimally separate important from negligible channels.

44, TITLE:      Neural CRF Model for Sentence Alignment in Text Simplification
http://arxiv.org/abs/2005.02324
AUTHORS:        Chao Jiang ; Mounica Maddela ; Wuwei Lan ; Yang Zhong ; Wei Xu
COMMENTS:       The paper has been accepted to ACL 2020
HIGHLIGHT:      We apply our CRF aligner to construct two new text simplification datasets, Newsela-Auto and Wiki-Auto, which are much larger and of better quality compared to the existing datasets. To evaluate and improve sentence alignment quality, we create two manually annotated sentence-aligned datasets from two commonly used text simplification corpora, Newsela and Wikipedia.

45, TITLE:      Understanding Generalization through Visualizations
http://arxiv.org/abs/1906.03291
AUTHORS:        W. Ronny Huang ; Zeyad Emam ; Micah Goldblum ; Liam Fowl ; Justin K. Terry ; Furong Huang ; Tom Goldstein
COMMENTS:       8 pages (excluding acknowledgments and references), 8 figures
HIGHLIGHT:      The goal of this work is to make generalization more intuitive.

46, TITLE:      Variance Reduction Methods for Sublinear Reinforcement Learning
http://arxiv.org/abs/1802.09184
AUTHORS:        Sham Kakade ; Mengdi Wang ; Lin F. Yang
COMMENTS:       There is a technical issue in the analysis that is not easily fixable
HIGHLIGHT:      Variance Reduction Methods for Sublinear Reinforcement Learning

47, TITLE:      A Methodology for Creating AI FactSheets
http://arxiv.org/abs/2006.13796
AUTHORS:        John Richards ; David Piorkowski ; Michael Hind ; Stephanie Houde ; Aleksandra Mojsilović
COMMENTS:       18 pages
HIGHLIGHT:      This paper describes this methodology and shares the insights we have gathered.

48, TITLE:      LAMP: Large Deep Nets with Automated Model Parallelism for Image Segmentation
http://arxiv.org/abs/2006.12575
AUTHORS:        Wentao Zhu ; Can Zhao ; Wenqi Li ; Holger Roth ; Ziyue Xu ; Daguang Xu
COMMENTS:       MICCAI 2020 Early Accepted paper. Code is available\footnote{https://monai.io/research/lamp-automated-model-parallelism}
HIGHLIGHT:      In this work, we introduce Large deep 3D ConvNets with Automated Model Parallelism (LAMP) and investigate the impact of both input's and deep 3D ConvNets' size on segmentation accuracy.

49, TITLE:      Modeling Label Semantics for Predicting Emotional Reactions
http://arxiv.org/abs/2006.05489
AUTHORS:        Radhika Gaonkar ; Heeyoung Kwon ; Mohaddeseh Bastan ; Niranjan Balasubramanian ; Nathanael Chambers
COMMENTS:       6 pages, 2 figures, published in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
HIGHLIGHT:      In this work, we explicitly model label classes via label embeddings, and add mechanisms that track label-label correlations both during training and inference.

50, TITLE:      BERT Loses Patience: Fast and Robust Inference with Early Exit
http://arxiv.org/abs/2006.04152
AUTHORS:        Wangchunshu Zhou ; Canwen Xu ; Tao Ge ; Julian McAuley ; Ke Xu ; Furu Wei
COMMENTS:       13 pages
HIGHLIGHT:      In this paper, we propose Patience-based Early Exit, a straightforward yet effective inference method that can be used as a plug-and-play technique to simultaneously improve the efficiency and robustness of a pretrained language model (PLM).

51, TITLE:      QUANOS- Adversarial Noise Sensitivity Driven Hybrid Quantization of Neural Networks
http://arxiv.org/abs/2004.11233
AUTHORS:        Priyadarshini Panda
COMMENTS:       Accepted in ACM/IEEE International Symposium on Low Power Electronics and Design (ISLPED), 2020
HIGHLIGHT:      In this paper, we investigate the use of quantization to potentially resist adversarial attacks.

52, TITLE:      The GraphNet Zoo: An All-in-One Graph Based Deep Semi-Supervised Framework for Medical Image Classification
http://arxiv.org/abs/2003.06451
AUTHORS:        Marianne de Vriendt ; Philip Sellars ; Angelica I Aviles-Rivero
HIGHLIGHT:      In this work, we propose an all-in-one framework for deep semi-supervised classification focusing on graph based approaches, which up to our knowledge it is the first time that an approach with minimal labels has been shown to such an unprecedented scale with medical data.

53, TITLE:      Limited View and Sparse Photoacoustic Tomography for Neuroimaging with Deep Learning
http://arxiv.org/abs/1911.04357
AUTHORS:        Steven Guan ; Amir A. Khan ; Siddhartha Sikdar ; Parag V. Chitnis
HIGHLIGHT:      In this work, we propose a novel deep learning approach termed pixelwise deep learning (PixelDL) that first employs pixelwise interpolation governed by the physics of photoacoustic wave propagation and then uses a convolution neural network to directly reconstruct an image.

54, TITLE:      Combining Deep Learning and Verification for Precise Object Instance Detection
http://arxiv.org/abs/1912.12270
AUTHORS:        Siddharth Ancha ; Junyu Nan ; David Held
COMMENTS:       9 pages main paper, 2 pages references, 10 pages supplementary material
HIGHLIGHT:      Based on an approximation to this framework, we present a practical detection system that can verify, with high precision, whether each detection of a machine-learning based object detector is correct. To achieve this, we have developed a set of verification tests which a proposed detection must pass to be accepted.

55, TITLE:      PowerNorm: Rethinking Batch Normalization in Transformers
http://arxiv.org/abs/2003.07845
AUTHORS:        Sheng Shen ; Zhewei Yao ; Amir Gholami ; Michael W. Mahoney ; Kurt Keutzer
HIGHLIGHT:      In this paper, we perform a systematic study of NLP transformer models to understand why BN has a poor performance, as compared to LN.

56, TITLE:      A multi-channel framework for joint reconstruction of multi-contrast parallel MRI
http://arxiv.org/abs/2006.04128
AUTHORS:        Erfan Ebrahim Esfahani
HIGHLIGHT:      In this paper, a novel isotropic multi-channel image regularizer is introduced, and its full potential is unleashed by integrating it into compressed multi-contrast multi-coil MRI.

57, TITLE:      Pure Pattern Calculus à la de Bruijn
http://arxiv.org/abs/2006.07674
AUTHORS:        Alexis Martín ; Alejandro Ríos ; Andrés Viso
HIGHLIGHT:      This paper extends de Bruijn's ideas to properly overcome the multi-binding problem by introducing a novel presentation of PPC with bidimensional indices, in an effort to implement a prototype for a typed functional programming language based on PPC that captures path polymorphism.

58, TITLE:      MILA: Multi-Task Learning from Videos via Efficient Inter-Frame Local Attention
http://arxiv.org/abs/2002.07362
AUTHORS:        Donghyun Kim ; Tian Lan ; Chuhang Zou ; Ning Xu ; Bryan A. Plummer ; Stan Sclaroff ; Jayan Eledath ; Gerard Medioni
HIGHLIGHT:      In this work, we present a new approach for multi-task learning from videos via efficient inter-frame local attention (MILA).

59, TITLE:      Tetra-Tagging: Word-Synchronous Parsing with Linear-Time Inference
http://arxiv.org/abs/1904.09745
AUTHORS:        Nikita Kitaev ; Dan Klein
COMMENTS:       ACL 2020
HIGHLIGHT:      We present a constituency parsing algorithm that, like a supertagger, works by assigning labels to each word in a sentence.

60, TITLE:      Knowledge of Uncertain Worlds: Programming with Logical Constraints
http://arxiv.org/abs/1910.10346
AUTHORS:        Yanhong A. Liu ; Scott D. Stoller
HIGHLIGHT:      This paper describes a unified language, DA logic, for design and analysis logic, based on the unifying founded semantics and constraint semantics, that support the power and ease of programming with different intended semantics.

61, TITLE:      Security and Privacy Preserving Deep Learning
http://arxiv.org/abs/2006.12698
AUTHORS:        Saichethan Miriyala Reddy ; Saisree Miriyala
HIGHLIGHT:      In this chapter, we introduce differential privacy, which ensures that different kinds of statistical analyses dont compromise privacy and federated learning, training a machine learning model on a data to which we do not have access to.

62, TITLE:      Affinity Fusion Graph-based Framework for Natural Image Segmentation
http://arxiv.org/abs/2006.13542
AUTHORS:        Yang Zhang ; Moyun Liu ; Jingwu He ; Fei Pan ; Yanwen Guo
COMMENTS:       11 pages, 10 figures
HIGHLIGHT:      This paper proposes an affinity fusion graph framework to effectively connect different graphs with highly discriminating power and nonlinearity for natural image segmentation.

63, TITLE:      Learning for Video Compression with Recurrent Auto-Encoder and Recurrent Probability Model
http://arxiv.org/abs/2006.13560
AUTHORS:        Ren Yang ; Fabian Mentzer ; Luc Van Gool ; Radu Timofte
COMMENTS:       This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
HIGHLIGHT:      To overcome this shortcoming, this paper proposes a Recurrent Learned Video Compression (RLVC) approach with the Recurrent Auto-Encoder (RAE) and Recurrent Probability Model (RPM).

64, TITLE:      ProSelfLC: Progressive Self Label Correction for Training Robust Deep Neural Networks
http://arxiv.org/abs/2005.03788
AUTHORS:        Xinshao Wang ; Yang Hua ; Elyor Kodirov ; Neil M. Robertson
COMMENTS:       Open discussion: we show it's fine for a learner (student) to be confident towards a correct low entropy status. Then more research attention should be paid to the definition of correct knowledge, as in genera we accept, human annotations used for learning supervision may be biased, subjective, and wrong
HIGHLIGHT:      To resolve these issues, taking two well-accepted expertise: deep neural networks learn meaningful patterns before fitting noise [1] and minimum entropy regularisation principle [2], we propose a simple end-to-end method named ProSelfLC, which is endorsed by long learning time and high prediction confidence.

65, TITLE:      Tensor Programs II: Neural Tangent Kernel for Any Architecture
http://arxiv.org/abs/2006.14548
AUTHORS:        Greg Yang
COMMENTS:       11 pages of main text. 60 pages total
HIGHLIGHT:      We prove that a randomly initialized neural network of *any architecture* has its Tangent Kernel (NTK) converge to a deterministic limit, as the network widths tend to infinity.

66, TITLE:      Querying and Repairing Inconsistent Prioritized Knowledge Bases: Complexity Analysis and Links with Abstract Argumentation
http://arxiv.org/abs/2003.05746
AUTHORS:        Meghyn Bienvenu ; Camille Bourgaux
COMMENTS:       27 pages. To appear in the 17th International Conference on Principles of Knowledge Representation and Reasoning (KR 2020) without the appendix
HIGHLIGHT:      In this paper, we explore the issue of inconsistency handling over prioritized knowledge bases (KBs), which consist of an ontology, a set of facts, and a priority relation between conflicting facts.

67, TITLE:      Compositional Learning of Image-Text Query for Image Retrieval
http://arxiv.org/abs/2006.11149
AUTHORS:        Muhammad Umer Anwaar ; Egor Labintcev ; Martin Kleinsteuber
HIGHLIGHT:      In this paper, we investigate the problem of retrieving images from a database based on a multi-modal (image-text) query.

68, TITLE:      GSTO: Gated Scale-Transfer Operation for Multi-Scale Feature Learning in Pixel Labeling
http://arxiv.org/abs/2005.13363
AUTHORS:        Zhuoying Wang ; Yongtao Wang ; Zhi Tang ; Yangyan Li ; Ying Chen ; Haibin Ling ; Weisi Lin
HIGHLIGHT:      In this work, we find that these operations lead to scale-confused features and suboptimal performance because they are spatial-invariant and directly transit all feature information cross scales without spatial selection.

69, TITLE:      Locally Masked Convolution for Autoregressive Models
http://arxiv.org/abs/2006.12486
AUTHORS:        Ajay Jain ; Pieter Abbeel ; Deepak Pathak
COMMENTS:       Published at Conference on Uncertainty in AI (UAI) 2020
HIGHLIGHT:      To generate data in arbitrary orders, we introduce LMConv: a simple modification to the standard 2D convolution that allows arbitrary masks to be applied to the weights at each location in the image.

70, TITLE:      Learning Heuristics over Large Graphs via Deep Reinforcement Learning
http://arxiv.org/abs/1903.03332
AUTHORS:        Sahil Manchanda ; Akash Mittal ; Anuj Dhawan ; Sourav Medya ; Sayan Ranu ; Ambuj Singh
HIGHLIGHT:      In this paper, we propose a framework called GCOMB to bridge these gaps.

71, TITLE:      Dueling Posterior Sampling for Preference-Based Reinforcement Learning
http://arxiv.org/abs/1908.01289
AUTHORS:        Ellen R. Novoseller ; Yibing Wei ; Yanan Sui ; Yisong Yue ; Joel W. Burdick
COMMENTS:       To appear in Conference on Uncertainty in Artificial Intelligence (UAI), 2020. 9 pages before references and appendix; 51 pages total; 7 figures; 4 tables. This replacement incorporates reviewer comments, and in comparison to version 1, extends the theoretical and empirical analyses and adds mathematical detail. Code: https://github.com/ernovoseller/DuelingPosteriorSampling
HIGHLIGHT:      Building upon ideas from preference-based bandit learning and posterior sampling in RL, we present DUELING POSTERIOR SAMPLING (DPS), which employs preference-based posterior sampling to learn both the system dynamics and the underlying utility function that governs the preference feedback.

72, TITLE:      Differential Dynamic Programming Neural Optimizer
http://arxiv.org/abs/2002.08809
AUTHORS:        Guan-Horng Liu ; Tianrong Chen ; Evangelos A. Theodorou
HIGHLIGHT:      In this work, we make an attempt along this line by reformulating the training procedure from the trajectory optimization perspective.

73, TITLE:      Privacy at Facebook Scale
http://arxiv.org/abs/2006.14109
AUTHORS:        Paulo Tanaka ; Sameet Sapra ; Nikolay Laptev
HIGHLIGHT:      Privacy at Facebook Scale

74, TITLE:      Self-supervised Label Augmentation via Input Transformations
http://arxiv.org/abs/1910.05872
AUTHORS:        Hankook Lee ; Sung Ju Hwang ; Jinwoo Shin
COMMENTS:       Accepted to ICML 2020. Code available at https://github.com/hankook/SLA
HIGHLIGHT:      In this paper, we show that such a technique can be used to significantly improve the model accuracy even under fully-labeled datasets.

75, TITLE:      Continual Representation Learning for Biometric Identification
http://arxiv.org/abs/2006.04455
AUTHORS:        Bo Zhao ; Shixiang Tang ; Dapeng Chen ; Hakan Bilen ; Rui Zhao
HIGHLIGHT:      In this paper, we propose a new continual learning (CL) setting, namely ``continual representation learning'', which focuses on learning better representation in a continuous way. We also provide two large-scale multi-step benchmarks for biometric identification, where the visual appearance of different classes are highly relevant.

76, TITLE:      A Tale of Two Perplexities: Sensitivity of Neural Language Models to Lexical Retrieval Deficits in Dementia of the Alzheimer's Type
http://arxiv.org/abs/2005.03593
AUTHORS:        Trevor Cohen ; Serguei Pakhomov
COMMENTS:       To be published in the Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL 2020)
HIGHLIGHT:      In this paper, we interrogate neural LMs trained on participants with and without dementia using synthetic narratives previously developed to simulate progressive semantic dementia by manipulating lexical frequency.

77, TITLE:      Filter Sketch for Network Pruning
http://arxiv.org/abs/2001.08514
AUTHORS:        Mingbao Lin ; Rongrong Ji ; Shaojie Li ; Qixiang Ye ; Yonghong Tian ; Jianzhuang Liu ; Qi Tian
HIGHLIGHT:      We propose a novel network pruning approach by information preserving of pre-trained network weights (filters).

78, TITLE:      Revisiting Agglomerative Clustering
http://arxiv.org/abs/2005.07995
AUTHORS:        Eric K. Tokuda ; Cesar H. Comin ; Luciano da F. Costa
HIGHLIGHT:      This work addressed this problem considering agglomerative methods, namely single, average, median, complete, centroid and Ward's approaches applied to unimodal and bimodal datasets obeying uniform, gaussian, exponential and power-law distributions.

79, TITLE:      Fairness-Aware Explainable Recommendation over Knowledge Graphs
http://arxiv.org/abs/2006.02046
AUTHORS:        Zuohui Fu ; Yikun Xian ; Ruoyuan Gao ; Jieyu Zhao ; Qiaoying Huang ; Yingqiang Ge ; Shuyuan Xu ; Shijie Geng ; Chirag Shah ; Yongfeng Zhang ; Gerard de Melo
HIGHLIGHT:      In this paper, we analyze different groups of users according to their level of activity, and find that bias exists in recommendation performance between different groups.

80, TITLE:      Inexpensive Domain Adaptation of Pretrained Language Models: Case Studies on Biomedical NER and Covid-19 QA
http://arxiv.org/abs/2004.03354
AUTHORS:        Nina Poerner ; Ulli Waltinger ; Hinrich Schütze
HIGHLIGHT:      We cover over 60% of the BioBERT-BERT F1 delta, at 5% of BioBERT's CO_2 footprint and 2% of its cloud compute cost.

81, TITLE:      DeepACEv2: Automated Chromosome Enumeration in Metaphase Cell Images Using Deep Convolutional Neural Networks
http://arxiv.org/abs/1910.11091
AUTHORS:        Chunlong Luo ; Li Xiao ; Tianqi Yu ; Yufan Luo ; Manqing Wang ; Fuhai Yu ; Yinhao Li ; Chan Tian ; Jie Qiao
HIGHLIGHT:      To automate the enumeration process, we develop a chromosome enumeration framework, DeepACEv2, based on the region based object detection scheme.

82, TITLE:      Exploring the ability of CNNs to generalise to previously unseen scales over wide scale ranges
http://arxiv.org/abs/2004.01536
AUTHORS:        Ylva Jansson ; Tony Lindeberg
COMMENTS:       Shortened version to appear in International Conference on Pattern Recognition (ICPR 2020). 14 pages, 6 figures
HIGHLIGHT:      We, therefore, present a theoretical analysis of invariance and covariance properties of scale channel networks and perform an experimental evaluation of the ability of different types of scale channel networks to generalise to previously unseen scales.

83, TITLE:      On Layer Normalization in the Transformer Architecture
http://arxiv.org/abs/2002.04745
AUTHORS:        Ruibin Xiong ; Yunchang Yang ; Di He ; Kai Zheng ; Shuxin Zheng ; Chen Xing ; Huishuai Zhang ; Yanyan Lan ; Liwei Wang ; Tie-Yan Liu
HIGHLIGHT:      Specifically, we prove with mean field theory that at initialization, for the original-designed Post-LN Transformer, which places the layer normalization between the residual blocks, the expected gradients of the parameters near the output layer are large.

84, TITLE:      Cartoon Face Recognition: A Benchmark Dataset
http://arxiv.org/abs/1907.13394
AUTHORS:        Yi Zheng ; Yifan Zhao ; Mengyuan Ren ; He Yan ; Xiangju Lu ; Junhui Liu ; Jia Li
COMMENTS:       9 papers, 6 figures
HIGHLIGHT:      To further investigate this challenging dataset, we propose a multi-task domain adaptation approach that jointly utilizes the human and cartoon domain knowledge with three discriminative regularizations.

85, TITLE:      Adaptive spline fitting with particle swarm optimization
http://arxiv.org/abs/1907.12160
AUTHORS:        Soumya D. Mohanty ; Ethan Fahnestock
COMMENTS:       Expanded literature survey; new figure added
HIGHLIGHT:      We present a method that uses particle swarm optimization (PSO) combined with model selection to address this challenge.

86, TITLE:      k-decay: A New Method For Learning Rate Schedule
http://arxiv.org/abs/2004.05909
AUTHORS:        Tao Zhang ; Wei Li
HIGHLIGHT:      In this paper, we propose the k-decay method, in which the LR is changed by its k-th order derivative, to obtain the new LR schedule.

87, TITLE:      Surrogate-assisted Bayesian inversion for landscape and basin evolution models
http://arxiv.org/abs/1812.08655
AUTHORS:        Rohitash Chandra ; Danial Azam ; Arpit Kapoor ; R. Dietmar Müller
COMMENTS:       Geoscientific Model Development
HIGHLIGHT:      In this paper, we apply surrogate-assisted parallel tempering where that surrogate mimics a landscape evolution model by estimating the likelihood function from the model.

88, TITLE:      Task-Agnostic Online Reinforcement Learning with an Infinite Mixture of Gaussian Processes
http://arxiv.org/abs/2006.11441
AUTHORS:        Mengdi Xu ; Wenhao Ding ; Jiacheng Zhu ; Zuxin Liu ; Baiming Chen ; Ding Zhao
COMMENTS:       16 pages, 6 figures
HIGHLIGHT:      This paper proposes a continual online model-based reinforcement learning approach that does not require pre-training to solve task-agnostic problems with unknown task boundaries.

89, TITLE:      Bayesian Reinforcement Learning via Deep, Sparse Sampling
http://arxiv.org/abs/1902.02661
AUTHORS:        Divya Grover ; Debabrota Basu ; Christos Dimitrakakis
COMMENTS:       Published in AISTATS 2020
HIGHLIGHT:      We propose an optimism-free Bayes-adaptive algorithm to induce deeper and sparser exploration with a theoretical bound on its performance relative to the Bayes optimal policy, with a lower computational complexity.

90, TITLE:      Pareto Optimization for Subset Selection with Dynamic Cost Constraints
http://arxiv.org/abs/1811.07806
AUTHORS:        Vahid Roostapour ; Aneta Neumann ; Frank Neumann ; Tobias Friedrich
COMMENTS:       A preliminary version of this article has been presented at the Thirty-Third AAAI Conference on Artificial Intelligence (AAAI 2019)
HIGHLIGHT:      We consider the subset selection problem for function $f$ with constraint bound $B$ that changes over time.

91, TITLE:      Visual Grounding of Learned Physical Models
http://arxiv.org/abs/2004.13664
AUTHORS:        Yunzhu Li ; Toru Lin ; Kexin Yi ; Daniel M. Bear ; Daniel L. K. Yamins ; Jiajun Wu ; Joshua B. Tenenbaum ; Antonio Torralba
COMMENTS:       The second and the third authors contributed equally to this paper, and are listed in alphabetical order. Project Page: http://visual-physics-grounding.csail.mit.edu/
HIGHLIGHT:      In this work, we present a neural model that simultaneously reasons about physics and makes future predictions based on visual and dynamics priors.

92, TITLE:      FGN: Fusion Glyph Network for Chinese Named Entity Recognition
http://arxiv.org/abs/2001.05272
AUTHORS:        Zhenyu Xuan ; Rui Bao ; Chuyu Ma ; Shengyi Jiang
HIGHLIGHT:      In this paper, we propose the FGN, Fusion Glyph Network for Chinese NER.

93, TITLE:      The Complexity of Verifying Loop-Free Programs as Differentially Private
http://arxiv.org/abs/1911.03272
AUTHORS:        Marco Gaboardi ; Kobbi Nissim ; David Purser
HIGHLIGHT:      We study the problem of verifying differential privacy for loop-free programs with probabilistic choice.

94, TITLE:      A Study on AI-FML Robotic Agent for Student Learning Behavior Ontology Construction
http://arxiv.org/abs/2006.10228
AUTHORS:        Chang-Shing Lee ; Mei-Hui Wang ; Wen-Kai Kuan ; Zong-Han Ciou ; Yi-Lin Tsai ; Wei-Shan Chang ; Lian-Chao Li ; Naoyuki Kubota ; Tzong-Xiang Huang ; Eri Sato-Shimokawara ; Toru Yamaguchi
COMMENTS:       This article has been accepted as a conference paper at CcS 2020 and will be published in IEEE
HIGHLIGHT:      In this paper, we propose an AI-FML robotic agent for student learning behavior ontology construction which can be applied in English speaking and listening domain.

95, TITLE:      To Ensemble or Not Ensemble: When does End-To-End Training Fail?
http://arxiv.org/abs/1902.04422
AUTHORS:        Andrew M. Webb ; Charles Reynolds ; Dan-Andrei Iliescu ; Henry Reeve ; Mikel Lujan ; Gavin Brown
COMMENTS:       Code: https://github.com/grey-area/modular-loss-experiments. Preprint updated to reflect version accepted for publication at ECML
HIGHLIGHT:      We study this question in depth, for the specific case of E2E training an ensemble of networks.

96, TITLE:      Predicting Livelihood Indicators from Crowdsourced Street Level Images
http://arxiv.org/abs/2006.08661
AUTHORS:        Jihyeon Janel Lee ; Dylan Grosz ; Sicheng Zeng ; Burak Uzkent ; Marshall Burke ; David Lobell ; Stefano Ermon
HIGHLIGHT:      We propose an inexpensive, scalable, and interpretable approach to predict key livelihood indicators from public crowd-sourced street-level imagery.

97, TITLE:      Encoding word order in complex embeddings
http://arxiv.org/abs/1912.12333
AUTHORS:        Benyou Wang ; Donghao Zhao ; Christina Lioma ; Qiuchi Li ; Peng Zhang ; Jakob Grue Simonsen
COMMENTS:       15 pages, 3 figures, ICLR 2020 spotlight paper. A typo on Ablation Table was revised thanks to Jingquan Zeng from SCUT
HIGHLIGHT:      We present a novel and principled solution for modeling both the global absolute positions of words and their order relationships.

98, TITLE:      Brain tumor segmentation with missing modalities via latent multi-source correlation representation
http://arxiv.org/abs/2003.08870
AUTHORS:        Tongxue Zhou ; Stéphane Canu ; Pierre Vera ; Su Ruan
COMMENTS:       9 pages, 6 figures, accepted by MICCAI 2020
HIGHLIGHT:      We evaluate our model on BraTS 2018 datasets, it outperforms the current state-of-the-art method and produces robust results when one or more modalities are missing.

99, TITLE:      PhishGAN: Data Augmentation and Identification of Homoglpyh Attacks
http://arxiv.org/abs/2006.13742
AUTHORS:        Joon Sern Lee ; Gui Peng David Yam ; Jin Hao Chan
COMMENTS:       8 pages, 8 figures
HIGHLIGHT:      Here, we show how a conditional Generative Adversarial Network (GAN), PhishGAN, can be used to generate images of hieroglyphs, conditioned on non-homoglpyh input text images.

100, TITLE:     Benchmarking Differentially Private Residual Networks for Medical Imagery
http://arxiv.org/abs/2005.13099
AUTHORS:        Sahib Singh ; Harshvardhan Sikka ; Sasikanth Kotti ; Andrew Trask
COMMENTS:       5 Pages, 3 Figures
HIGHLIGHT:      In this paper, we measure the performance of a deep neural network on differentially private image datasets pertaining to Pneumonia.

101, TITLE:     Linking Points With Labels in 3D: A Review of Point Cloud Semantic Segmentation
http://arxiv.org/abs/1908.08854
AUTHORS:        Yuxing Xie ; Jiaojiao Tian ; Xiao Xiang Zhu
COMMENTS:       The title of published version was modified to "Linking Points With Labels in 3D: A Review of Point Cloud Semantic Segmentation". To read its final version please go to IEEE Geoscience and Remote Sensing Magazine on IEEE XPlore: https://ieeexplore.ieee.org/document/9028090
HIGHLIGHT:      In order to provide a needed up-to-date review of recent developments in PCSS, this article summarizes existing studies on this topic.

102, TITLE:     Assessment Modeling: Fundamental Pre-training Tasks for Interactive Educational Systems
http://arxiv.org/abs/2002.05505
AUTHORS:        Youngduck Choi ; Youngnam Lee ; Junghyun Cho ; Jineon Baek ; Dongmin Shin ; Seewoo Lee ; Jonghun Shin ; Chan Bae ; Byungsoo Kim ; Jaewe Heo
HIGHLIGHT:      To this end, we propose assessment modeling, fundamental pre-training tasks for IESs.

103, TITLE:     Classification of Epithelial Ovarian Carcinoma Whole-Slide Pathology Images Using Deep Transfer Learning
http://arxiv.org/abs/2005.10957
AUTHORS:        Yiping Wang ; David Farnell ; Hossein Farahani ; Mitchell Nursey ; Basile Tessier-Cloutier ; Steven J. M. Jones ; David G. Huntsman ; C. Blake Gilks ; Ali Bashashati
HIGHLIGHT:      We utilized a \textit{two}-stage deep transfer learning algorithm based on convolutional neural networks (CNN) and progressive resizing for automatic classification of epithelial ovarian carcinoma WSIs.

104, TITLE:     Recurrent Hierarchical Topic-Guided RNN for Language Generation
http://arxiv.org/abs/1912.10337
AUTHORS:        Dandan Guo ; Bo Chen ; Ruiying Lu ; Mingyuan Zhou
COMMENTS:       ICML 2020
HIGHLIGHT:      To simultaneously capture syntax and global semantics from a text corpus, we propose a new larger-context recurrent neural network (RNN) based language model, which extracts recurrent hierarchical semantic structure via a dynamic deep topic model to guide natural language generation.

105, TITLE:     PointAR: Efficient Lighting Estimation for Mobile Augmented Reality
http://arxiv.org/abs/2004.00006
AUTHORS:        Yiqin Zhao ; Tian Guo
HIGHLIGHT:      We propose an efficient lighting estimation pipeline that is suitable to run on modern mobile devices, with comparable resource complexities to state-of-the-art on-device deep learning models.

106, TITLE:     Deep Encoder, Shallow Decoder: Reevaluating the Speed-Quality Tradeoff in Machine Translation
http://arxiv.org/abs/2006.10369
AUTHORS:        Jungo Kasai ; Nikolaos Pappas ; Hao Peng ; James Cross ; Noah A. Smith
COMMENTS:       Added references and acknowledgements
HIGHLIGHT:      In this work, we re-examine the trade-off and argue that transformer-based autoregressive models can be substantially sped up without loss in accuracy.

107, TITLE:     Learning image registration without images
http://arxiv.org/abs/2004.10282
AUTHORS:        Malte Hoffmann ; Benjamin Billot ; Juan Eugenio Iglesias ; Bruce Fischl ; Adrian V. Dalca
COMMENTS:       17 pages, 12 figures, deformable image registration, contrast invariance, training without data, expanded analyses
HIGHLIGHT:      We introduce a learning strategy for contrast-invariant image registration without requiring imaging data.

108, TITLE:     Graph Optimal Transport for Cross-Domain Alignment
http://arxiv.org/abs/2006.14744
AUTHORS:        Liqun Chen ; Zhe Gan ; Yu Cheng ; Linjie Li ; Lawrence Carin ; Jingjing Liu
HIGHLIGHT:      We propose Graph Optimal Transport (GOT), a principled framework that germinates from recent advances in Optimal Transport (OT).

109, TITLE:     Implicit Pairs for Boosting Unpaired Image-to-Image Translation
http://arxiv.org/abs/1904.06913
AUTHORS:        Yiftach Ginger ; Dov Danon ; Hadar Averbuch-Elor ; Daniel Cohen-Or
HIGHLIGHT:      In our work, we show that injecting implicit pairs into unpaired sets strengthens the mapping between the two domains, improves the compatibility of their distributions, and leads to performance boosting of unsupervised techniques by over 14% across several measurements.

110, TITLE:     Rule Extraction in Unsupervised Anomaly Detection for Model Explainability: Application to OneClass SVM
http://arxiv.org/abs/1911.09315
AUTHORS:        Alberto Barbado ; Óscar Corcho ; Richard Benjamins
COMMENTS:       23 pages, 18 figures
HIGHLIGHT:      Together with that, we propose algorithms to compute metrics related with XAI regarding the "comprehensivility", "representativeness", "stability" and "diversity" of the rules extracted.

111, TITLE:     Stance in Replies and Quotes (SRQ): A New Dataset For Learning Stance in Twitter Conversations
http://arxiv.org/abs/2006.00691
AUTHORS:        Ramon Villa-Cox ; Sumeet Kumar ; Matthew Babcock ; Kathleen M. Carley
HIGHLIGHT:      In the presented work, we create a new dataset by labeling stance in responses to posts on Twitter (both replies and quotes) on controversial issues.

112, TITLE:     Exploiting Contextual Information with Deep Neural Networks
http://arxiv.org/abs/2006.11706
AUTHORS:        Ismail Elezi
COMMENTS:       Ph.D. thesis
HIGHLIGHT:      In this thesis, we show that contextual information can be exploited in 2 fundamentally different ways: implicitly and explicitly.

113, TITLE:     Fine-tuning Multi-hop Question Answering with Hierarchical Graph Network
http://arxiv.org/abs/2004.13821
AUTHORS:        Guanming Xiong
COMMENTS:       the experience result is not as good as I except
HIGHLIGHT:      In this paper, we present a two stage model for multi-hop question answering.

114, TITLE:     Weak Supervision and Referring Attention for Temporal-Textual Association Learning
http://arxiv.org/abs/2006.11747
AUTHORS:        Zhiyuan Fang ; Shu Kong ; Zhe Wang ; Charless Fowlkes ; Yezhou Yang
COMMENTS:       12 pages, 6 figures
HIGHLIGHT:      Therefore we provide a Weak-Supervised alternative with our proposed Referring Attention mechanism to learn temporal-textual association (dubbed WSRA).

115, TITLE:     Viral Pneumonia Screening on Chest X-ray Images Using Confidence-Aware Anomaly Detection
http://arxiv.org/abs/2003.12338
AUTHORS:        Jianpeng Zhang ; Yutong Xie ; Zhibin Liao ; Guansong Pang ; Johan Verjans ; Wenxin Li ; Zongji Sun ; Jian He ; Yi Li ; Chunhua Shen ; Yong Xia
HIGHLIGHT:      In this paper, we formulate the task of differentiating viral pneumonia from non-viral pneumonia and healthy controls into an one-class classification-based anomaly detection problem, and thus propose the confidence-aware anomaly detection (CAAD) model, which consists of a shared feature extractor, an anomaly detection module, and a confidence prediction module.

116, TITLE:     Learning to Dynamically Coordinate Multi-Robot Teams in Graph Attention Networks
http://arxiv.org/abs/1912.02059
AUTHORS:        Zheyuan Wang ; Matthew Gombolay
COMMENTS:       This paper has been extended to an article in IEEE Robotics and Automation Letters (DOI: 10.1109/LRA.2020.3002198)
HIGHLIGHT:      To learn effective policies for combinatorial optimization problems via machine learning, we combine imitation learning on smaller problems with deep Q-learning on larger problems, in a non-parametric framework, to allow for fast, near-optimal scheduling of robot teams.

117, TITLE:     Intelligent Coordination among Multiple Traffic Intersections Using Multi-Agent Reinforcement Learning
http://arxiv.org/abs/1912.03851
AUTHORS:        Ujwal Padam Tewari ; Vishal Bidawatka ; Varsha Raveendran ; Vinay Sudhakaran ; Shreedhar Kodate Shreeshail ; Jayanth Prakash Kulkarni
COMMENTS:       Accepted in the NeurIPS 2019 Deep RL Workshop : https://sites.google.com/view/deep-rl-workshop-neurips-2019/home
HIGHLIGHT:      We use Asynchronous Advantage Actor Critic (A3C) for implementing an AI agent in the controllers that optimize flow of traffic across a single intersection and then extend it to multiple intersections by considering a multi-agent setting.

118, TITLE:     Predicting Temporal Sets with Deep Neural Networks
http://arxiv.org/abs/2006.11483
AUTHORS:        Le Yu ; Leilei Sun ; Bowen Du ; Chuanren Liu ; Hui Xiong ; Weifeng Lv
COMMENTS:       9 pages, 6 figures, Proceedings of the 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '2020)
HIGHLIGHT:      In this paper, we propose an integrated solution based on the deep neural networks for temporal sets prediction.

119, TITLE:     Real-time Fusion Network for RGB-D Semantic Segmentation Incorporating Unexpected Obstacle Detection for Road-driving Images
http://arxiv.org/abs/2002.10570
AUTHORS:        Lei Sun ; Kailun Yang ; Xinxin Hu ; Weijian Hu ; Kaiwei Wang
COMMENTS:       Accepted by IEEE Robotics and Automation Letters (RA-L); 8 Figures, 3 Tables; Code is available at https://github.com/AHupuJR/RFNet
HIGHLIGHT:      In this paper, we propose a real-time fusion semantic segmentation network termed RFNet that effectively exploits complementary cross-modal information.

120, TITLE:     AdvAug: Robust Adversarial Augmentation for Neural Machine Translation
http://arxiv.org/abs/2006.11834
AUTHORS:        Yong Cheng ; Lu Jiang ; Wolfgang Macherey ; Jacob Eisenstein
COMMENTS:       fix the typo
HIGHLIGHT:      In this paper, we propose a new adversarial augmentation method for Neural Machine Translation (NMT).

121, TITLE:     Object-Oriented Video Captioning with Trajectory Graph and Attribute Exploring
http://arxiv.org/abs/2003.03715
AUTHORS:        Fangyi Zhu ; Jenq-Neng Hwang ; Zhanyu Ma ; Guang Chen ; Jun Guo
HIGHLIGHT:      We propose a novel task, named object-oriented video captioning, focusing on understanding the videos in object-level. Thereafter, we construct a new dataset with explicit object-sentence pairs to facilitate effective cross-modal learning.

122, TITLE:     3D Object Detection on Point Clouds using Local Ground-aware and Adaptive Representation of scenes' surface
http://arxiv.org/abs/2002.00336
AUTHORS:        Arun CS Kumar ; Disha Ahuja ; Ashwath Aithal
HIGHLIGHT:      A novel, adaptive ground-aware, and cost-effective 3D Object Detection pipeline is proposed.

123, TITLE:     Self-Supervised Learning for Domain Adaptation on Point-Clouds
http://arxiv.org/abs/2003.12641
AUTHORS:        Idan Achituve ; Haggai Maron ; Gal Chechik
HIGHLIGHT:      We introduce a new family of pretext tasks, \textit{Deformation Reconstruction}, motivated by the deformations encountered in sim-to-real transformations.

124, TITLE:     Automatic Estimation of Intelligibility Measure for Consonants in Speech
http://arxiv.org/abs/2005.06065
AUTHORS:        Ali Abavisani ; Mark Hasegawa-Johnson
COMMENTS:       5 pages, 1 figure, 7 tables, submitted to Inter Speech 2020 Conference
HIGHLIGHT:      In this article, we provide a model to estimate a real-valued measure of the intelligibility of individual speech segments.