2020.06.11.txt

==========New Papers==========
1, TITLE:       Predicting and Analyzing Law-Making in Kenya
http://arxiv.org/abs/2006.05493
AUTHORS:        Oyinlola Babafemi ; Adewale Akinfaderin
COMMENTS:       Accepted at 4th Widening NLP Workshop, Annual Meeting of the Association for Computational Linguistics, ACL 2020
HIGHLIGHT:      In this paper, we focused on understanding the bills introduced in a developing democracy, the Kenyan bicameral parliament.

2, TITLE:       Deep Learning with Attention Mechanism for Predicting Driver Intention at Intersection
http://arxiv.org/abs/2006.05918
AUTHORS:        Abenezer Girma ; Seifemichael Amsalu ; Abrham Workineh ; Mubbashar Khan ; Abdollah Homaifar
COMMENTS:       IEEE Intelligent Vehicles Symposium 2020 (IEEE IV 2020)
HIGHLIGHT:      In this paper, a driver's intention prediction near a road intersection is proposed.

3, TITLE:       Embed2Detect: Temporally Clustered Embedded Words for Event Detection in Social Media
http://arxiv.org/abs/2006.05908
AUTHORS:        Hansi Hettiarachchi ; Mariam Adedoyin-Olowe ; Jagdev Bhogal ; Mohamed Medhat Gaber
COMMENTS:       Submitted to Journal of Neural Computing and Applications, Springer
HIGHLIGHT:      In this paper, we propose a novel method termed Embed2Detect for event detection in social media by combining the characteristics in prediction-based word embeddings and hierarchical agglomerative clustering.

4, TITLE:       Data Augmentation for Training Dialog Models Robust to Speech Recognition Errors
http://arxiv.org/abs/2006.05635
AUTHORS:        Longshaokan Wang ; Maryam Fazel-Zarandi ; Aditya Tiwari ; Spyros Matsoukas ; Lazaros Polymenakos
COMMENTS:       To be presented at 2nd Workshop on NLP for ConvAI, ACL 2020
HIGHLIGHT:      To bridge the gap and make dialog models more robust to ASR errors, we leverage an ASR error simulator to inject noise into the error-free text data, and subsequently train the dialog models with the augmented data.

5, TITLE:       Retrofitting Symbolic Holes to LLVM IR
http://arxiv.org/abs/2006.05875
AUTHORS:        Bruce Collie ; Michael O'Boyle
COMMENTS:       Accepted to TyDe 2020
HIGHLIGHT:      This paper motivates and introduces the implementation of symbolic holes with unknown type to LLVM IR, a strongly-typed compiler intermediate language.

6, TITLE:       WasteNet: Waste Classification at the Edge for Smart Bins
http://arxiv.org/abs/2006.05873
AUTHORS:        Gary White ; Christian Cabrera ; Andrei Palade ; Fan Li ; Siobhan Clarke
COMMENTS:       8 pages, 9 figures
HIGHLIGHT:      In this paper, we propose WasteNet, a waste classification model based on convolutional neural networks that can be deployed on a low power device at the edge of the network, such as a Jetson Nano.

7, TITLE:       Improving Dependability of Neuromorphic Computing With Non-Volatile Memory
http://arxiv.org/abs/2006.05868
AUTHORS:        Shihao Song ; Anup Das ; Nagarajan Kandasamy
COMMENTS:       8 pages, 13 figures, accepted in 16th European Dependable Computing Conference
HIGHLIGHT:      This paper proposes RENEU, a reliability-oriented approach to map machine learning applications to neuromorphic hardware, with the aim of improving system-wide reliability without compromising key performance metrics such as execution time of these applications on the hardware.

8, TITLE:       A survey on deep hashing for image retrieval
http://arxiv.org/abs/2006.05627
AUTHORS:        Xiaopeng Zhang
HIGHLIGHT:      In this survey, several deep supervised hashing methods for image retrieval are evaluated and I conclude three main different directions for deep supervised hashing methods.

9, TITLE:       Better Together: Resnet-50 accuracy with $13x$ fewer parameters and at $3x$ speed
http://arxiv.org/abs/2006.05624
AUTHORS:        Utkarsh Nath ; Shrinu Kushagra
COMMENTS:       Code available at: https://github.com/utkarshnath/Adjoint-Network.git
HIGHLIGHT:      We introduce Adjoined networks as a training approach that can compress and regularize any CNN-based neural architecture.

10, TITLE:      Objective Caml for Multicore Architectures
http://arxiv.org/abs/2006.05862
AUTHORS:        Mathias Bourgoin ; Benjamin Canou ; Emmanuel Chailloux ; Adrien Jonquet ; Philippe Wang
HIGHLIGHT:      This paper presents our feedback on removing Objective Caml's garbage collector and designing a "Stop-The-World Stop&Copy" garbage collector to permit threads to take advantage of multicore architectures.

11, TITLE:      Understanding Points of Correspondence between Sentences for Abstractive Summarization
http://arxiv.org/abs/2006.05621
AUTHORS:        Logan Lebanoff ; John Muchovej ; Franck Dernoncourt ; Doo Soon Kim ; Lidan Wang ; Walter Chang ; Fei Liu
COMMENTS:       Camera-ready version for ACL 2020 Student Research Workshop (SRW)
HIGHLIGHT:      In this paper, we present an investigation into fusing sentences drawn from a document by introducing the notion of points of correspondence, which are cohesive devices that tie any two sentences together into a coherent text. We create a dataset containing the documents, source and fusion sentences, and human annotations of points of correspondence between sentences.

12, TITLE:      A systematic review on the role of artificial intelligence in sonographic diagnosis of thyroid cancer: Past, present and future
http://arxiv.org/abs/2006.05861
AUTHORS:        Fatemeh Abdolali ; Atefeh Shahroudnejad ; Abhilash Rakkunedeth Hareendranathan ; Jacob L Jaremko ; Michelle Noga ; Kumaradevan Punithakumar
HIGHLIGHT:      In this paper, we present a systematic review of state-of-the-art on AI application in sonographic diagnosis of thyroid cancer.

13, TITLE:      Deep Learning for Change Detection in Remote Sensing Images: Comprehensive Review and Meta-Analysis
http://arxiv.org/abs/2006.05612
AUTHORS:        Lazhar Khelifi ; Max Mignotte
HIGHLIGHT:      This study will contribute in several ways to our understanding of deep learning for change detection and will provide a basis for further research.

14, TITLE:      Geometry-Aware Segmentation of Remote Sensing Images via implicit height estimation
http://arxiv.org/abs/2006.05848
AUTHORS:        Xiang Li ; Yi Fang
COMMENTS:       13 pages, 10 figures
HIGHLIGHT:      To alleviate this data constraint and also take the advantage of 3D elevation information, in this paper, we propose a geometry-aware segmentation model that achieves accurate semantic segmentation of aerial images via implicit height estimation.

15, TITLE:      Toward a standardized methodology for constructing quantum computing use cases
http://arxiv.org/abs/2006.05846
AUTHORS:        Nicholas Chancellor ; Robert Cumming ; Tim Thomas
COMMENTS:       9 pages, 3 figures
HIGHLIGHT:      The purpose of this paper is to initiate a dialogue within the community of quantum computing scientists and potential end users on what questions should be asked when developing real world use cases.

16, TITLE:      Searching Learning Strategy with Reinforcement Learning for 3D Medical Image Segmentation
http://arxiv.org/abs/2006.05847
AUTHORS:        Dong Yang ; Holger Roth ; Ziyue Xu ; Fausto Milletari ; Ling Zhang ; Daguang Xu
COMMENTS:       9 pages, 1 figures
HIGHLIGHT:      Nevertheless, to fully exploit the potentials of neural networks, we propose an automated searching approach for the optimal training strategy with reinforcement learning.

17, TITLE:      Adversarial Training Based Multi-Source Unsupervised Domain Adaptation for Sentiment Analysis
http://arxiv.org/abs/2006.05602
AUTHORS:        Yong Dai ; Jian Liu ; Xiancong Ren ; Zenglin Xu
HIGHLIGHT:      To avoid these problems, we propose two transfer learning frameworks based on the multi-source domain adaptation methodology for SA by combining the source hypotheses to derive a good target hypothesis.

18, TITLE:      The Emergence of Individuality in Multi-Agent Reinforcement Learning
http://arxiv.org/abs/2006.05842
AUTHORS:        Jiechuan Jiang ; Zongqing Lu
HIGHLIGHT:      Inspired by that individuality is of being an individual separate from others, we propose a simple yet efficient method for the emergence of individuality (EOI) in multi-agent reinforcement learning (MARL).

19, TITLE:      Simultaneous Decision Making for Stochastic Multi-echelon Inventory Optimization with Deep Neural Networks as Decision Makers
http://arxiv.org/abs/2006.05608
AUTHORS:        Mohammad Pirhooshyaran ; Lawrence V. Snyder
HIGHLIGHT:      We present a framework which uses deep neural networks as agents responsible for finding order-up-to levels for any desired components of the general supply chain network.

20, TITLE:      To Regularize or Not To Regularize? The Bias Variance Trade-off in Regularized AEs
http://arxiv.org/abs/2006.05838
AUTHORS:        Arnab Kumar Mondal ; Himanshu Asnani ; Parag Singla ; Prathosh AP
HIGHLIGHT:      Motivated by this, we examine the effect of the latent prior on the generation quality of the AE models in this paper.

21, TITLE:      The Impact of Non-stationarity on Generalisation in Deep Reinforcement Learning
http://arxiv.org/abs/2006.05826
AUTHORS:        Maximilian Igl ; Gregory Farquhar ; Jelena Luketina ; Wendelin Boehmer ; Shimon Whiteson
HIGHLIGHT:      Consequently, to improve generalisation of deep RL agents, we propose Iterated Relearning (ITER).

22, TITLE:      Development of A Stochastic Traffic Environment with Generative Time-Series Models for Improving Generalization Capabilities of Autonomous Driving Agents
http://arxiv.org/abs/2006.05821
AUTHORS:        Anil Ozturk ; Mustafa Burak Gunel ; Melih Dal ; Ugur Yavas ; Nazim Kemal Ure
COMMENTS:       7 pages, 4 figures, 7 tables, IV2020
HIGHLIGHT:      In this work, we develop a data driven traffic simulator by training a generative adverserial network (GAN) on real life trajectory data.

23, TITLE:      DcardNet: Diabetic Retinopathy Classification at Multiple Depths Based on Structural and Angiographic Optical Coherence Tomography
http://arxiv.org/abs/2006.05480
AUTHORS:        Pengxiao Zang ; Liqin Gao ; Tristan T. Hormel ; Jie Wang ; Qisheng You ; Thomas S. Hwang ; Yali Jia
COMMENTS:       Submitted to IEEE Transactions on Biomedical Engineering
HIGHLIGHT:      In this study, a densely and continuously connected neural network with adaptive rate dropout (DcardNet) is proposed to fulfill a DR classification framework using en face OCT and OCTA.

24, 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, to be published in The 58th Annual Meeting of the Association for Computational Linguistics 2020
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.

25, TITLE:      Off-the-shelf sensor vs. experimental radar -- How much resolution is necessary in automotive radar classification?
http://arxiv.org/abs/2006.05485
AUTHORS:        Nicolas Scheiner ; Ole Schumann ; Florian Kraus ; Nils Appenrodt ; Jürgen Dickmann ; Bernhard Sick
COMMENTS:       Accepted @ 23rd International Conference on Information Fusion (FUSION)
HIGHLIGHT:      In this article, two sensors of different radar generations are evaluated against each other.

26, TITLE:      Standardised convolutional filtering for radiomics
http://arxiv.org/abs/2006.05470
AUTHORS:        Adrien Depeursinge ; Vincent Andrearczyk ; Philip Whybra ; Joost van Griethuysen ; Henning Müller ; Roger Schaer ; Martin Vallières ; Alex Zwanenburg
COMMENTS:       54 pages. For additional information see https://theibsi.github.io/
HIGHLIGHT:      Here we present a preliminary version of a reference manual on the use of convolutional image filters in radiomics.

27, TITLE:      Unsupervised Paraphrase Generation using Pre-trained Language Models
http://arxiv.org/abs/2006.05477
AUTHORS:        Chaitra Hegde ; Shrikumar Patil
HIGHLIGHT:      In this paper we leverage this generation capability of GPT-2 to generate paraphrases without any supervision from labelled data.

28, TITLE:      Improving Cross-Lingual Transfer Learning for End-to-End Speech Recognition with Speech Translation
http://arxiv.org/abs/2006.05474
AUTHORS:        Changhan Wang ; Juan Pino ; Jiatao Gu
COMMENTS:       Submitted to INTERSPEECH 2020
HIGHLIGHT:      We introduce speech-to-text translation (ST) as an auxiliary task to incorporate additional knowledge of the target language and enable transferring from that target language.

29, TITLE:      Examination and Extension of Strategies for Improving Personalized Language Modeling via Interpolation
http://arxiv.org/abs/2006.05469
AUTHORS:        Liqun Shao ; Sahitya Mantravadi ; Tom Manzini ; Alejandro Buendia ; Manon Knoertzer ; Soundar Srinivasan ; Chris Quirk
COMMENTS:       ACL Natural Language Interface Workshop 2020, short paper
HIGHLIGHT:      In this paper, we detail novel strategies for interpolating personalized language models and methods to handle out-of-vocabulary (OOV) tokens to improve personalized language models.

30, TITLE:      Pruning neural networks without any data by iteratively conserving synaptic flow
http://arxiv.org/abs/2006.05467
AUTHORS:        Hidenori Tanaka ; Daniel Kunin ; Daniel L. K. Yamins ; Surya Ganguli
HIGHLIGHT:      This raises a foundational question: can we identify highly sparse trainable subnetworks at initialization, without ever training, or indeed without ever looking at the data?

31, TITLE:      Rendering Natural Camera Bokeh Effect with Deep Learning
http://arxiv.org/abs/2006.05698
AUTHORS:        Andrey Ignatov ; Jagruti Patel ; Radu Timofte
HIGHLIGHT:      Unlike the current solutions simulating bokeh by applying Gaussian blur to image background, in this paper we propose to learn a realistic shallow focus technique directly from the photos produced by DSLR cameras. For this, we present a large-scale bokeh dataset consisting of 5K shallow / wide depth-of-field image pairs captured using the Canon 7D DSLR with 50mm f/1.8 lenses.

32, TITLE:      Dialog Policy Learning for Joint Clarification and Active Learning Queries
http://arxiv.org/abs/2006.05456
AUTHORS:        Aishwarya Padmakumar ; Raymond J. Mooney
HIGHLIGHT:      In this work, we train a hierarchical dialog policy to jointly perform {\it both} clarification and active learning in the context of an interactive language-based image retrieval task motivated by an on-line shopping application, and demonstrate that jointly learning dialog policies for clarification and active learning is more effective than the use of static dialog policies for one or both of these functions.

33, TITLE:      Variational Model-based Policy Optimization
http://arxiv.org/abs/2006.05443
AUTHORS:        Yinlam Chow ; Brandon Cui ; MoonKyung Ryu ; Mohammad Ghavamzadeh
HIGHLIGHT:      In this paper, we leverage the connection between RL and probabilistic inference, and formulate such an objective function as a variational lower-bound of a log-likelihood.

34, TITLE:      H3DNet: 3D Object Detection Using Hybrid Geometric Primitives
http://arxiv.org/abs/2006.05682
AUTHORS:        Zaiwei Zhang ; Bo Sun ; Haitao Yang ; Qixing Huang
HIGHLIGHT:      We introduce H3DNet, which takes a colorless 3D point cloud as input and outputs a collection of oriented object bounding boxes (or BB) and their semantic labels.

35, TITLE:      TubeTK: Adopting Tubes to Track Multi-Object in a One-Step Training Model
http://arxiv.org/abs/2006.05683
AUTHORS:        Bo Pang ; Yizhuo Li ; Yifan Zhang ; Muchen Li ; Cewu Lu
COMMENTS:       CVPR-2020 oral paper
HIGHLIGHT:      To address these challenges, we propose a concise end-to-end model TubeTK which only needs one step training by introducing the ``bounding-tube" to indicate temporal-spatial locations of objects in a short video clip.

36, TITLE:      OpEvo: An Evolutionary Method for Tensor Operator Optimization
http://arxiv.org/abs/2006.05664
AUTHORS:        Xiaotian Gao ; Cui Wei ; Lintao Zhang ; Mao Yang
HIGHLIGHT:      In this work, we propose a novel evolutionary method, OpEvo, which efficiently explores the search spaces of tensor operators by introducing a topology-aware mutation operation based on q-random walk distribution to leverage the topological structures over the search spaces.

37, TITLE:      Methodology for Realizing VMM with Binary RRAM Arrays: Experimental Demonstration of Binarized-ADALINE Using OxRAM Crossbar
http://arxiv.org/abs/2006.05657
AUTHORS:        Sandeep Kaur Kingra ; Vivek Parmar ; Shubham Negi ; Sufyan Khan ; Boris Hudec ; Tuo-Hung Hou ; Manan Suri
COMMENTS:       Accepted for presentation at the IEEE International Symposium on Circuits and Systems (ISCAS) 2020
HIGHLIGHT:      In this paper, we present an efficient hardware mapping methodology for realizing vector matrix multiplication (VMM) on resistive memory (RRAM) arrays.

38, TITLE:      DisCont: Self-Supervised Visual Attribute Disentanglement using Context Vectors
http://arxiv.org/abs/2006.05895
AUTHORS:        Sarthak Bhagat ; Vishaal Udandarao ; Shagun Uppal
COMMENTS:       10 pages, 6 figures, 4 tables
HIGHLIGHT:      In this paper, we propose a self-supervised framework DisCont to disentangle multiple attributes by exploiting the structural inductive biases within images.

39, TITLE:      Rinascimento: using event-value functions for playing Splendor
http://arxiv.org/abs/2006.05894
AUTHORS:        Ivan Bravi ; Simon Lucas
COMMENTS:       To appear in IEEE Conference on Games 2019 Proceedings
HIGHLIGHT:      This paper proposes a new approach based on event logging: the game state triggers an event every time one of its features changes.

40, TITLE:      Agrupamento de Pixels para o Reconhecimento de Faces
http://arxiv.org/abs/2006.05652
AUTHORS:        Tiago Buarque Assunção de Carvalho
COMMENTS:       21 pages, in Portuguese, 5 figures, book chapter, recortado (adapatado) da tese
HIGHLIGHT:      To explain this fact, we proposed the Pixel Clustering methodology.

41, TITLE:      Tight Quantum Time-Space Tradeoffs for Function Inversion
http://arxiv.org/abs/2006.05650
AUTHORS:        Kai-Min Chung ; Siyao Guo ; Qipeng Liu ; Luowen Qian
HIGHLIGHT:      In this work, we prove that even with quantum advice, $ST + T^2 = \tilde\Omega(N)$ is required for an algorithm to invert random functions.

42, TITLE:      Speech Fusion to Face: Bridging the Gap Between Human's Vocal Characteristics and Facial Imaging
http://arxiv.org/abs/2006.05888
AUTHORS:        Yeqi Bai ; Tao Ma ; Lipo Wang ; Zhenjie Zhang
HIGHLIGHT:      In this paper, we investigate these key technical challenges and propose Speech Fusion to Face, or SF2F in short, attempting to address the issue of facial image quality and the poor connection between vocal feature domain and modern image generation models.

43, TITLE:      Scalable Backdoor Detection in Neural Networks
http://arxiv.org/abs/2006.05646
AUTHORS:        Haripriya Harikumar ; Vuong Le ; Santu Rana ; Sourangshu Bhattacharya ; Sunil Gupta ; Svetha Venkatesh
HIGHLIGHT:      In this paper, we propose a novel trigger reverse-engineering based approach whose computational complexity does not scale with the number of labels, and is based on a measure that is both interpretable and universal across different network and patch types.

44, TITLE:      Benchmarking a $(μ+λ)$ Genetic Algorithm with Configurable Crossover Probability
http://arxiv.org/abs/2006.05889
AUTHORS:        Furong Ye ; Hao Wang ; Carola Doerr ; Thomas Bäck
HIGHLIGHT:      We investigate a family of $(\mu+\lambda)$ Genetic Algorithms (GAs) which creates offspring either from mutation or by recombining two randomly chosen parents.

45, TITLE:      Supervised Learning of Sparsity-Promoting Regularizers for Denoising
http://arxiv.org/abs/2006.05521
AUTHORS:        Michael T. McCann ; Saiprasad Ravishankar
HIGHLIGHT:      We present a method for supervised learning of sparsity-promoting regularizers for image denoising.

46, TITLE:      MVLidarNet: Real-Time Multi-Class Scene Understanding for Autonomous Driving Using Multiple Views
http://arxiv.org/abs/2006.05518
AUTHORS:        Ke Chen ; Ryan Oldja ; Nikolai Smolyanskiy ; Stan Birchfield ; Alexander Popov ; David Wehr ; Ibrahim Eden ; Joachim Pehserl
COMMENTS:       IROS2020 conference submission, for accompanying video, see https://youtu.be/2ck5_sToayc
HIGHLIGHT:      To this end, we present a two-stage deep neural network (MVLidarNet) for multi-class object detection and drivable segmentation using multiple views of a single LiDAR point cloud.

47, TITLE:      Gender in Danger? Evaluating Speech Translation Technology on the MuST-SHE Corpus
http://arxiv.org/abs/2006.05754
AUTHORS:        Luisa Bentivogli ; Beatrice Savoldi ; Matteo Negri ; Mattia Antonino Di Gangi ; Roldano Cattoni ; Marco Turchi
COMMENTS:       9 pages of content, accepted at ACL 2020
HIGHLIGHT:      We present the first thorough investigation of gender bias in speech translation, contributing with: i) the release of a benchmark useful for future studies, and ii) the comparison of different technologies (cascade and end-to-end) on two language directions (English-Italian/French).

48, TITLE:      A Deep Learning-Based Method for Automatic Segmentation of Proximal Femur from Quantitative Computed Tomography Images
http://arxiv.org/abs/2006.05513
AUTHORS:        Chen Zhao ; Joyce H. Keyak ; Jinshan Tang ; Tadashi S. Kaneko ; Sundeep Khosla ; Shreyasee Amin ; Elizabeth J. Atkinson ; Lan-Juan Zhao ; Michael J. Serou ; Chaoyang Zhang ; Hui Shen ; Hong-Wen Deng ; Weihua Zhou
HIGHLIGHT:      We aim to develop a deep-learning-based method for automatic proximal femur segmentation.

49, TITLE:      Revisiting Few-sample BERT Fine-tuning
http://arxiv.org/abs/2006.05987
AUTHORS:        Tianyi Zhang ; Felix Wu ; Arzoo Katiyar ; Kilian Q. Weinberger ; Yoav Artzi
HIGHLIGHT:      We study the problem of few-sample fine-tuning of BERT contextual representations, and identify three sub-optimal choices in current, broadly adopted practices.

50, TITLE:      ClarQ: A large-scale and diverse dataset for Clarification Question Generation
http://arxiv.org/abs/2006.05986
AUTHORS:        Vaibhav Kumar ; Alan W. black
COMMENTS:       Accepted at ACL 2020
HIGHLIGHT:      In order to overcome these limitations, we devise a novel bootstrapping framework (based on self-supervision) that assists in the creation of a diverse, large-scale dataset of clarification questions based on post-comment tuples extracted from stackexchange. We release this dataset in order to foster research into the field of clarification question generation with the larger goal of enhancing dialog and question answering systems.

51, TITLE:      MC-BERT: Efficient Language Pre-Training via a Meta Controller
http://arxiv.org/abs/2006.05744
AUTHORS:        Zhenhui Xu ; Linyuan Gong ; Guolin Ke ; Di He ; Shuxin Zheng ; Liwei Wang ; Jiang Bian ; Tie-Yan Liu
HIGHLIGHT:      To achieve better efficiency and effectiveness, we propose a novel meta-learning framework, MC-BERT.

52, TITLE:      Can artificial intelligence (AI) be used to accurately detect tuberculosis (TB) from chest x-ray? A multiplatform evaluation of five AI products used for TB screening in a high TB-burden setting
http://arxiv.org/abs/2006.05509
AUTHORS:        Zhi Zhen Qin ; Shahriar Ahmed ; Mohammad Shahnewaz Sarker ; Kishor Paul ; Ahammad Shafiq Sikder Adel ; Tasneem Naheyan ; Sayera Banu ; Jacob Creswell
COMMENTS:       27 pages, 5 Tables 3 Figures
HIGHLIGHT:      The 23,566 individuals included in the study all received a CXR read by a group of three Bangladeshi board-certified radiologists.

53, TITLE:      3D Human Mesh Regression with Dense Correspondence
http://arxiv.org/abs/2006.05734
AUTHORS:        Wang Zeng ; Wanli Ouyang ; Ping Luo ; Wentao Liu ; Xiaogang Wang
COMMENTS:       To appear at CVPR 2020
HIGHLIGHT:      This paper proposes a model-free 3D human mesh estimation framework, named DecoMR, which explicitly establishes the dense correspondence between the mesh and the local image features in the UV space (i.e. a 2D space used for texture mapping of 3D mesh).

54, TITLE:      Object Detection in the DCT Domain: is Luminance the Solution?
http://arxiv.org/abs/2006.05732
AUTHORS:        Benjamin Deguerre ; Clement Chatelain ; Gilles Gasso
HIGHLIGHT:      To gain in efficiency, this paper proposes to take advantage of the compressed representation of images to carry out object detection usable in constrained resources conditions.

55, TITLE:      Diagnosing Rarity in Human-Object Interaction Detection
http://arxiv.org/abs/2006.05728
AUTHORS:        Mert Kilickaya ; Arnold Smeulders
COMMENTS:       Accepted at CVPR'20 Workshop on Learning from Limited Labels
HIGHLIGHT:      To that end, in this paper, we propose to diagnose rarity in HOI detection.

56, TITLE:      Bayesian Experience Reuse for Learning from Multiple Demonstrators
http://arxiv.org/abs/2006.05725
AUTHORS:        Michael Gimelfarb ; Scott Sanner ; Chi-Guhn Lee
COMMENTS:       15 pages, 7 figures
HIGHLIGHT:      We address this problem in the static and dynamic optimization settings by modelling the uncertainty in source and target task functions using normal-inverse-gamma priors, whose corresponding posteriors are, respectively, learned from demonstrations and target data using Bayesian neural networks with shared features.

57, TITLE:      Estimating semantic structure for the VQA answer space
http://arxiv.org/abs/2006.05726
AUTHORS:        Corentin Kervadec ; Grigory Antipov ; Moez Baccouche ; Christian Wolf
HIGHLIGHT:      We address this issue by proposing (1) two measures of proximity between VQA classes, and (2) a corresponding loss which takes into account the estimated proximity.

58, TITLE:      Real-time single image depth perception in the wild with handheld devices
http://arxiv.org/abs/2006.05724
AUTHORS:        Filippo Aleotti ; Giulio Zaccaroni ; Luca Bartolomei ; Matteo Poggi ; Fabio Tosi ; Stefano Mattoccia
COMMENTS:       11 pages, 9 figures
HIGHLIGHT:      Therefore, in this paper, we deeply investigate these issues showing how they are both addressable adopting appropriate network design and training strategies -- also outlining how to map the resulting networks on handheld devices to achieve real-time performance.

59, TITLE:      Balancing Fairness and Efficiency in an Optimization Model
http://arxiv.org/abs/2006.05963
AUTHORS:        Violet Xinying Chen ; J. N. Hooker
HIGHLIGHT:      We propose a principled and practical method for balancing these two criteria in an optimization model.

60, TITLE:      A Bayesian Framework for Nash Equilibrium Inference in Human-Robot Parallel Play
http://arxiv.org/abs/2006.05729
AUTHORS:        Shray Bansal ; Jin Xu ; Ayanna Howard ; Charles Isbell
COMMENTS:       Accepted at Robotics: Science and Systems (RSS) 2020
HIGHLIGHT:      We model these as general-sum games and construct a framework that utilizes the Nash equilibrium solution concept to consider the interactive effect of both agents while planning.

61, TITLE:      At-Most-One Constraints in Efficient Representations of Mutex Networks
http://arxiv.org/abs/2006.05962
AUTHORS:        Pavel Surynek
HIGHLIGHT:      An on-line method for automated detection of cliques for efficient representation of incremental mutex networks where new mutexes arrive using AMOs is presented.

62, TITLE:      OptiLIME: Optimized LIME Explanations for Diagnostic Computer Algorithms
http://arxiv.org/abs/2006.05714
AUTHORS:        Giorgio Visani ; Enrico Bagli ; Federico Chesani
HIGHLIGHT:      In this paper, we highlight a trade-off between explanation's stability and adherence, namely how much it resembles the ML model.

63, TITLE:      Unique Faces Recognition in Videos
http://arxiv.org/abs/2006.05713
AUTHORS:        Jiahao Huo ; Terence L van Zyl
COMMENTS:       Paper was accepted into Fusion 2020 conference but will only be published after the virtual conference in July 2020. 7 pages long
HIGHLIGHT:      The contribution of this paper is two-fold: to begin, the experiments have established 3-D Convolutional networks and 2-D LSTMs with the contrastive loss on image sequences do not outperform Google/Inception architecture with contrastive loss in top $n$ rank face retrievals with still images.

64, TITLE:      Few-shot Slot Tagging with Collapsed Dependency Transfer and Label-enhanced Task-adaptive Projection Network
http://arxiv.org/abs/2006.05702
AUTHORS:        Yutai Hou ; Wanxiang Che ; Yongkui Lai ; Zhihan Zhou ; Yijia Liu ; Han Liu ; Ting Liu
COMMENTS:       Accepted by ACL2020
HIGHLIGHT:      In this paper, we explore the slot tagging with only a few labeled support sentences (a.k.a. few-shot).

65, TITLE:      MultiResolution Attention Extractor for Small Object Detection
http://arxiv.org/abs/2006.05941
AUTHORS:        Fan Zhang ; Licheng Jiao ; Lingling Li ; Fang Liu ; Xu Liu
COMMENTS:       11 pages, 5 figures
HIGHLIGHT:      Inspired by human vision "attention" mechanism, we exploit two feature extraction methods to mine the most useful information of small objects.

66, TITLE:      Delta Descriptors: Change-Based Place Representation for Robust Visual Localization
http://arxiv.org/abs/2006.05700
AUTHORS:        Sourav Garg ; Ben Harwood ; Gaurangi Anand ; Michael Milford
COMMENTS:       8 pages and 7 figures. To be published in 2020 IEEE Robotics and Automation Letters (RA-L)
HIGHLIGHT:      In this paper we propose a novel descriptor derived from tracking changes in any learned global descriptor over time, dubbed Delta Descriptors.

67, TITLE:      Simple and effective localized attribute representations for zero-shot learning
http://arxiv.org/abs/2006.05938
AUTHORS:        Shiqi Yang ; Kai Wang ; Luis Herranz ; Joost van de Weijer
HIGHLIGHT:      In contrast, in this paper we propose localizing representations in the semantic/attribute space, with a simple but effective pipeline where localization is implicit.

68, TITLE:      Separable Four Points Fundamental Matrix
http://arxiv.org/abs/2006.05926
AUTHORS:        Gil Ben-Artzi
HIGHLIGHT:      We present an approach for the computation of the fundamental matrix based on epipolar homography decomposition.

69, TITLE:      Dataset Condensation with Gradient Matching
http://arxiv.org/abs/2006.05929
AUTHORS:        Bo Zhao ; Konda Reddy Mopuri ; Hakan Bilen
HIGHLIGHT:      This paper proposes a training set synthesis technique, called Dataset Condensation, that learns to produce a small set of informative samples for training deep neural networks from scratch in a small fraction of the required computational cost on the original data while achieving comparable results.

70, TITLE:      Recent Advances in 3D Object and Hand Pose Estimation
http://arxiv.org/abs/2006.05927
AUTHORS:        Vincent Lepetit
HIGHLIGHT:      In this chapter, we present the recent developments for 3D object and hand pose estimation using cameras, and discuss their abilities and limitations and the possible future development of the field.

71, TITLE:      Condensing Two-stage Detection with Automatic Object Key Part Discovery
http://arxiv.org/abs/2006.05597
AUTHORS:        Zhe Chen ; Jing Zhang ; Dacheng Tao
HIGHLIGHT:      To address this problem, we propose that the model parameters of two-stage detection heads can be condensed and reduced by concentrating on object key parts.

72, TITLE:      Fitted Q-Learning for Relational Domains
http://arxiv.org/abs/2006.05595
AUTHORS:        Srijita Das ; Sriraam Natarajan ; Kaushik Roy ; Ronald Parr ; Kristian Kersting
COMMENTS:       10 pages, 12 figures
HIGHLIGHT:      We consider the problem of Approximate Dynamic Programming in relational domains.

73, TITLE:      CNN-Based Semantic Change Detection in Satellite Imagery
http://arxiv.org/abs/2006.05589
AUTHORS:        Ananya Gupta ; Elisabeth Welburn ; Simon Watson ; Hujun Yin
HIGHLIGHT:      Herein, we propose a CNN-based framework for identifying accessible roads in post-disaster imagery by detecting changes from pre-disaster imagery.

74, TITLE:      Syn2Real Transfer Learning for Image Deraining using Gaussian Processes
http://arxiv.org/abs/2006.05580
AUTHORS:        Rajeev Yasarla Vishwanath A. Sindagi Vishal M. Patel
COMMENTS:       Accepted at CVPR 2020
HIGHLIGHT:      We propose a Gaussian Process-based semi-supervised learning framework which enables the network in learning to derain using synthetic dataset while generalizing better using unlabeled real-world images.

75, TITLE:      Dual-level Semantic Transfer Deep Hashing for Efficient Social Image Retrieval
http://arxiv.org/abs/2006.05586
AUTHORS:        Lei Zhu ; Hui Cui ; Zhiyong Cheng ; Jingjing Li ; Zheng Zhang
COMMENTS:       Accepted by IEEE TCSVT
HIGHLIGHT:      Differently, in this paper, we propose a Dual-level Semantic Transfer Deep Hashing (DSTDH) method to alleviate this problem with a unified deep hash learning framework.

76, TITLE:      Deep Learning-based Aerial Image Segmentation with Open Data for Disaster Impact Assessment
http://arxiv.org/abs/2006.05575
AUTHORS:        Ananya Gupta ; Simon Watson ; Hujun Yin
COMMENTS:       Accepted in Neurocomputing, 2020
HIGHLIGHT:      In order to provide timely and actionable information for disaster response, in this paper a framework utilising segmentation neural networks is proposed to identify impacted areas and accessible roads in post-disaster scenarios.

77, TITLE:      A gaze driven fast-forward method for first-person videos
http://arxiv.org/abs/2006.05569
AUTHORS:        Alan Carvalho Neves ; Michel Melo Silva ; Mario Fernando Montenegro Campos ; Erickson Rangel Nascimento
COMMENTS:       Accepted for presentation at EPIC@CVPR2020 workshop
HIGHLIGHT:      In this paper, we address the problem of accessing relevant information in First-Person Videos by creating an accelerated version of the input video and emphasizing the important moments to the recorder.

78, TITLE:      Learning Functions to Study the Benefit of Multitask Learning
http://arxiv.org/abs/2006.05561
AUTHORS:        Gabriele Bettgenhäuser ; Michael A. Hedderich ; Dietrich Klakow
HIGHLIGHT:      To remedy these limitations, we propose the creation of a task simulator and the use of Symbolic Regression to learn expressions relating model performance to possible factors of influence.

79, TITLE:      Tree Annotations in LiDAR Data Using Point Densities and Convolutional Neural Networks
http://arxiv.org/abs/2006.05560
AUTHORS:        Ananya Gupta ; Jonathan Byrne ; David Moloney ; Simon Watson ; Hujun Yin
HIGHLIGHT:      Manual annotation of such data is time consuming, tedious and error prone, and hence in this paper we present three automatic methods for annotating trees in LiDAR data.

80, TITLE:      Embedding Task Knowledge into 3D Neural Networks via Self-supervised Learning
http://arxiv.org/abs/2006.05798
AUTHORS:        Jiuwen Zhu ; Yuexiang Li ; Yifan Hu ; S. Kevin Zhou
HIGHLIGHT:      In this paper, we propose a novel SSL approach for 3D medical image classification, namely Task-related Contrastive Prediction Coding (TCPC), which embeds task knowledge into training 3D neural networks.

81, TITLE:      3D Point Cloud Feature Explanations Using Gradient-Based Methods
http://arxiv.org/abs/2006.05548
AUTHORS:        Ananya Gupta ; Simon Watson ; Hujun Yin
COMMENTS:       Accepted for IJCNN 2020
HIGHLIGHT:      We extend the saliency methods that have been shown to work on image data to deal with 3D data.

82, TITLE:      Image Enhancement and Object Recognition for Night Vision Surveillance
http://arxiv.org/abs/2006.05787
AUTHORS:        Aashish Bhandari ; Aayush Kafle ; Pranjal Dhakal ; Prateek Raj Joshi ; Dinesh Baniya Kshatri
COMMENTS:       International Conference on Recent Trends in Computational Engineering and Technologies, 2018
HIGHLIGHT:      The accuracy of classification after implementing different enhancement algorithms is compared in this paper.

83, TITLE:      Resolution-Enhanced MRI-Guided Navigation of Spinal Cellular Injection Robot
http://arxiv.org/abs/2006.05544
AUTHORS:        Daniel Enrique Martinez ; Waiman Meinhold ; John Oshinski ; Ai-Ping Hu ; Jun Ueda
COMMENTS:       6 pages, 10 figures, 3 tables, conference
HIGHLIGHT:      This paper presents a method of navigating a surgical robot beyond the resolution of magnetic resonance imaging (MRI) by using a resolution enhancement technique enabled by high-precision piezoelectric actuation.

84, TITLE:      Applying Deep-Learning-Based Computer Vision to Wireless Communications: Methodologies, Opportunities, and Challenges
http://arxiv.org/abs/2006.05782
AUTHORS:        Yu Tian ; Gaofeng Pan ; Mohamed-Slim Alouini
HIGHLIGHT:      In this example, we proposed a framework to predict the future beam indices from the previously-observed beam indices and images of street views by using ResNet, 3-dimensional ResNext, and long short term memory network.

85, TITLE:      Dual-stream Maximum Self-attention Multi-instance Learning
http://arxiv.org/abs/2006.05538
AUTHORS:        Bin Li ; Kevin W. Eliceiri
HIGHLIGHT:      In this paper, we proposed a dual-stream maximum self-attention MIL model (DSMIL) parameterized by neural networks.

86, TITLE:      Self-Supervised Reinforcement Learning forRecommender Systems
http://arxiv.org/abs/2006.05779
AUTHORS:        Xin Xin ; Alexandros Karatzoglou ; Ioannis Arapakis ; Joemon M. Jose
COMMENTS:       SIGIR2020
HIGHLIGHT:      In this paper, we propose self-supervised reinforcement learning for sequential recommendation tasks.



==========Updates to Previous Papers==========
1, TITLE:       Approximate Inverse Reinforcement Learning from Vision-based Imitation Learning
http://arxiv.org/abs/2004.08051
AUTHORS:        Keuntaek Lee ; Bogdan Vlahov ; Jason Gibson ; James M. Rehg ; Evangelos A. Theodorou
HIGHLIGHT:      In this work, we present a method for obtaining an implicit objective function for vision-based navigation.

2, TITLE:       Exact algorithms for semidefinite programs with degenerate feasible set
http://arxiv.org/abs/1802.02834
AUTHORS:        Didier Henrion ; Simone Naldi ; Mohab Safey El Din
COMMENTS:       26 pages, 1 figure, extended version (the original paper is published in the Proceedings of ISSAC 2018)
HIGHLIGHT:      In this paper, we design an exact algorithm based on symbolic homotopy for solving semidefinite programs without assumptions on the feasible set, and we analyze its complexity.

3, TITLE:       Two-stage dimension reduction for noisy high-dimensional images and application to Cryogenic Electron Microscopy
http://arxiv.org/abs/1911.09816
AUTHORS:        Szu-Chi Chung ; Shao-Hsuan Wang ; Po-Yao Niu ; Su-Yun Huang ; Wei-Hau Chang ; I-Ping Tu
COMMENTS:       29 pages, 8 figures and 3 tables
HIGHLIGHT:      We propose herein a two-stage dimension reduction (2SDR) method for image reconstruction from high-dimensional noisy image data.

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

5, TITLE:       Self-Paced Deep Regression Forests with Consideration on Underrepresented Samples
http://arxiv.org/abs/2004.01459
AUTHORS:        Lili Pan ; Shijie Ai ; Yazhou Ren ; Zenglin Xu
COMMENTS:       18 pages, 6 figures
HIGHLIGHT:      To this end, this paper proposes a new deep discriminative model -- self-paced deep regression forests with consideration on underrepresented samples (SPUDRFs).

6, TITLE:       3D Photography using Context-aware Layered Depth Inpainting
http://arxiv.org/abs/2004.04727
AUTHORS:        Meng-Li Shih ; Shih-Yang Su ; Johannes Kopf ; Jia-Bin Huang
COMMENTS:       CVPR 2020. Project page: https://shihmengli.github.io/3D-Photo-Inpainting/ Code: https://github.com/vt-vl-lab/3d-photo-inpainting Demo: https://colab.research.google.com/drive/1706ToQrkIZshRSJSHvZ1RuCiM__YX3Bz
HIGHLIGHT:      We propose a method for converting a single RGB-D input image into a 3D photo - a multi-layer representation for novel view synthesis that contains hallucinated color and depth structures in regions occluded in the original view.

7, TITLE:       VecQ: Minimal Loss DNN Model Compression With Vectorized Weight Quantization
http://arxiv.org/abs/2005.08501
AUTHORS:        Cheng Gong ; Yao Chen ; Ye Lu ; Tao Li ; Cong Hao ; Deming Chen
COMMENTS:       14 pages, 9 figures, Journal
HIGHLIGHT:      In this paper, we propose a novel metric called Vector Loss.

8, TITLE:       Text Processing Like Humans Do: Visually Attacking and Shielding NLP Systems
http://arxiv.org/abs/1903.11508
AUTHORS:        Steffen Eger ; Gözde Gül Şahin ; Andreas Rücklé ; Ji-Ung Lee ; Claudia Schulz ; Mohsen Mesgar ; Krishnkant Swarnkar ; Edwin Simpson ; Iryna Gurevych
COMMENTS:       Accepted as long paper at NAACL-2019; fixed one ungrammatical sentence
HIGHLIGHT:      We consider this as a new type of adversarial attack in NLP, a setting to which humans are very robust, as our experiments with both simple and more difficult visual input perturbations demonstrate.

9, TITLE:       Predictive Coding Approximates Backprop along Arbitrary Computation Graphs
http://arxiv.org/abs/2006.04182
AUTHORS:        Beren Millidge ; Alexander Tschantz ; Christopher L. Buckley
COMMENTS:       Submitted to NeurIPS 2020. Updated Acknowledgements
HIGHLIGHT:      Here, we demonstrate that predictive coding converges asymptotically (and in practice rapidly) to exact backprop gradients on arbitrary computation graphs using only local learning rules.

10, TITLE:      Measuring Diversity of Artificial Intelligence Conferences
http://arxiv.org/abs/2001.07038
AUTHORS:        Ana Freire ; Lorenzo Porcaro ; Emilia Gómez
HIGHLIGHT:      We consider diversity in terms of gender, geographical location and business (understood as the presence of academia versus industry).

11, TITLE:      Perturbed Masking: Parameter-free Probing for Analyzing and Interpreting BERT
http://arxiv.org/abs/2004.14786
AUTHORS:        Zhiyong Wu ; Yun Chen ; Ben Kao ; Qun Liu
COMMENTS:       ACL2020
HIGHLIGHT:      Complementary to those works, we propose a parameter-free probing technique for analyzing pre-trained language models (e.g., BERT).

12, TITLE:      Adaptive Transformers for Learning Multimodal Representations
http://arxiv.org/abs/2005.07486
AUTHORS:        Prajjwal Bhargava
COMMENTS:       Accepted at ACL SRW 2020. Code can be found here https://github.com/prajjwal1/adaptive_transformer
HIGHLIGHT:      In this work, we extend adaptive approaches to learn more about model interpretability and computational efficiency.

13, TITLE:      AutoGrow: Automatic Layer Growing in Deep Convolutional Networks
http://arxiv.org/abs/1906.02909
AUTHORS:        Wei Wen ; Feng Yan ; Yiran Chen ; Hai Li
COMMENTS:       KDD 2020
HIGHLIGHT:      We propose AutoGrow to automate depth discovery in DNNs: starting from a shallow seed architecture, AutoGrow grows new layers if the growth improves the accuracy; otherwise, stops growing and thus discovers the depth.

14, TITLE:      CoVoST: A Diverse Multilingual Speech-To-Text Translation Corpus
http://arxiv.org/abs/2002.01320
AUTHORS:        Changhan Wang ; Juan Pino ; Anne Wu ; Jiatao Gu
COMMENTS:       LREC 2020 camera-ready
HIGHLIGHT:      We introduce CoVoST, a multilingual speech-to-text translation corpus from 11 languages into English, diversified with over 11,000 speakers and over 60 accents. We also provide initial benchmarks, including, to our knowledge, the first end-to-end many-to-one multilingual models for spoken language translation. We also provide additional evaluation data derived from Tatoeba under CC licenses.

15, TITLE:      Training End-to-End Analog Neural Networks with Equilibrium Propagation
http://arxiv.org/abs/2006.01981
AUTHORS:        Jack Kendall ; Ross Pantone ; Kalpana Manickavasagam ; Yoshua Bengio ; Benjamin Scellier
HIGHLIGHT:      We introduce a principled method to train end-to-end analog neural networks by stochastic gradient descent.

16, TITLE:      Compressing deep neural networks by matrix product operators
http://arxiv.org/abs/1904.06194
AUTHORS:        Ze-Feng Gao ; Song Cheng ; Rong-Qiang He ; Z. Y. Xie ; Hui-Hai Zhao ; Zhong-Yi Lu ; Tao Xiang
COMMENTS:       8+9 pages, 3+7 figures, 2+11 tables
HIGHLIGHT:      Here we show that this problem can be effectively solved by representing linear transformations with matrix product operators (MPOs), which is a tensor network originally proposed in physics to characterize the short-range entanglement in one-dimensional quantum states.

17, TITLE:      Alleviating Semantic-level Shift: A Semi-supervised Domain Adaptation Method for Semantic Segmentation
http://arxiv.org/abs/2004.00794
AUTHORS:        Zhonghao Wang ; Yunchao Wei ; Rogerior Feris ; Jinjun Xiong ; Wen-Mei Hwu ; Thomas S. Huang ; Humphrey Shi
COMMENTS:       CVPRW 2020
HIGHLIGHT:      To tackle this issue, we propose a semi-supervised approach named Alleviating Semantic-level Shift (ASS), which can successfully promote the distribution consistency from both global and local views.

18, TITLE:      Policy-Driven Neural Response Generation for Knowledge-Grounded Dialogue Systems
http://arxiv.org/abs/2005.12529
AUTHORS:        Behnam Hedayatnia ; Karthik Gopalakrishnan ; Seokhwan Kim ; Yang Liu ; Mihail Eric ; Dilek Hakkani-Tur
COMMENTS:       Typos in Figure 2 and 6 Typo in author listing
HIGHLIGHT:      In this paper, we propose using a dialogue policy to plan the content and style of target responses in the form of an action plan, which includes knowledge sentences related to the dialogue context, targeted dialogue acts, topic information, etc.

19, TITLE:      'Target Set Selection' on Graphs of Bounded Vertex Cover Number
http://arxiv.org/abs/1812.01482
AUTHORS:        Suman Banerjee ; Rogers Mathew ; Fahad Panolan
COMMENTS:       11 pages
HIGHLIGHT:      Given a simple, undirected graph $G$ with a threshold function $\tau:V(G) \rightarrow \mathbb{N}$, the \textsc{Target Set Selection} (TSS) Problem is about choosing a minimum cardinality set, say $S \subseteq V(G)$, such that starting a diffusion process with $S$ as its seed set will eventually result in activating all the nodes in $G$.

20, TITLE:      Off-policy Bandit and Reinforcement Learning
http://arxiv.org/abs/2002.08536
AUTHORS:        Yusuke Narita ; Shota Yasui ; Kohei Yata
HIGHLIGHT:      We develop a method for predicting the performance of reinforcement learning and bandit algorithms, given historical data that may have been generated by a different algorithm.

21, TITLE:      Language Modeling for Formal Mathematics
http://arxiv.org/abs/2006.04757
AUTHORS:        Markus N. Rabe ; Dennis Lee ; Kshitij Bansal ; Christian Szegedy
HIGHLIGHT:      To train language models for formal mathematics, we propose a novel skip-tree task, which outperforms standard language modeling tasks on our reasoning benchmarks.

22, TITLE:      Exploration of Interpretability Techniques for Deep COVID-19 Classification using Chest X-ray Images
http://arxiv.org/abs/2006.02570
AUTHORS:        Soumick Chatterjee ; Fatima Saad ; Chompunuch Sarasaen ; Suhita Ghosh ; Rupali Khatun ; Petia Radeva ; Georg Rose ; Sebastian Stober ; Oliver Speck ; Andreas Nürnberger
HIGHLIGHT:      Thereby, the use of five different deep learning models (ResNet18, ResNet34, InceptionV3, InceptionResNetV2, and DenseNet161) and their Ensemble have been used in this paper, to classify COVID-19, pneumoni{\ae} and healthy subjects using Chest X-Ray.

23, TITLE:      On equivalence, languages equivalence and minimization of multi-letter and multi-letter measure-many quantum automata
http://arxiv.org/abs/1203.0113
AUTHORS:        Tianrong Lin
COMMENTS:       30 pages
HIGHLIGHT:      The direct consequences of the above outcomes are summarized in the paper.

24, TITLE:      Automatic Translating Between Ancient Chinese and Contemporary Chinese with Limited Aligned Corpora
http://arxiv.org/abs/1803.01557
AUTHORS:        Zhiyuan Zhang ; Wei Li ; Qi Su
COMMENTS:       Acceptted by NLPCC 2019
HIGHLIGHT:      In this paper, we propose to build an end-to-end neural model to automatically translate between ancient and contemporary Chinese.

25, 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.

26, TITLE:      Position-based Scaled Gradient for Model Quantization and Sparse Training
http://arxiv.org/abs/2005.11035
AUTHORS:        Jangho Kim ; KiYoon Yoo ; Nojun Kwak
HIGHLIGHT:      We propose the position-based scaled gradient (PSG) that scales the gradient depending on the position of a weight vector to make it more compression-friendly.

27, TITLE:      Interpretable Random Forests via Rule Extraction
http://arxiv.org/abs/2004.14841
AUTHORS:        Clément Bénard ; Gérard Biau ; Sébastien da Veiga ; Erwan Scornet
HIGHLIGHT:      We introduce SIRUS (Stable and Interpretable RUle Set) for regression, a stable rule learning algorithm which takes the form of a short and simple list of rules.

28, TITLE:      Non-congruent non-degenerate curves with identical signatures
http://arxiv.org/abs/1912.09597
AUTHORS:        Eric Geiger ; Irina A. Kogan
COMMENTS:       32 pages, 26 figures. Updated to include results on local symmetries and improvements to presentation
HIGHLIGHT:      A solution to this problem based on the differential signatures was proposed by Calabi et al. (Int.

29, TITLE:      Learning to Reach Goals via Iterated Supervised Learning
http://arxiv.org/abs/1912.06088
AUTHORS:        Dibya Ghosh ; Abhishek Gupta ; Ashwin Reddy ; Justin Fu ; Coline Devin ; Benjamin Eysenbach ; Sergey Levine
COMMENTS:       First two authors contributed equally. Code available at https://github.com/notdibya/gcsl
HIGHLIGHT:      In this paper, we study RL algorithms for learning goal reaching policies that leverage the stability of imitation learning without the need for explicit expert demonstrations.

30, TITLE:      Instance Scale Normalization for image understanding
http://arxiv.org/abs/1908.07323
AUTHORS:        Zewen He ; He Huang ; Yudong Wu ; Guan Huang ; Wensheng Zhang
HIGHLIGHT:      In this work, we propose an innovative paradigm called Instance Scale Normalization (ISN) to resolve the above problem.

31, TITLE:      Fixed-Support Wasserstein Barycenters: Computational Hardness and Fast Algorithm
http://arxiv.org/abs/2002.04783
AUTHORS:        Tianyi Lin ; Nhat Ho ; Xi Chen ; Marco Cuturi ; Michael I. Jordan
COMMENTS:       Change the title; Improve writing and correct some typos
HIGHLIGHT:      We study the fixed-support Wasserstein barycenter problem (FS-WBP), which consists in computing the Wasserstein barycenter of $m$ discrete probability measures supported on a finite metric space of size $n$.

32, TITLE:      On Model Robustness Against Adversarial Examples
http://arxiv.org/abs/1911.06479
AUTHORS:        Shufei Zhang ; Kaizhu Huang ; Zenglin Xu
COMMENTS:       some theoretical bounds need to be revised
HIGHLIGHT:      We propose to exploit an energy function to describe the stability and prove that reducing such energy guarantees the robustness against adversarial examples.

33, TITLE:      Predicting Engagement in Video Lectures
http://arxiv.org/abs/2006.00592
AUTHORS:        Sahan Bulathwela ; María Pérez-Ortiz ; Aldo Lipani ; Emine Yilmaz ; John Shawe-Taylor
COMMENTS:       In Proceedings of International Conference on Educational Data Mining 2020
HIGHLIGHT:      In this work, we explore the idea of building a predictive model for population-based engagement in education. We introduce a novel, large dataset of video lectures for predicting context-agnostic engagement and propose both cross-modal and modality-specific feature sets to achieve this task.

34, TITLE:      Learning Generative Models using Denoising Density Estimators
http://arxiv.org/abs/2001.02728
AUTHORS:        Siavash A. Bigdeli ; Geng Lin ; Tiziano Portenier ; L. Andrea Dunbar ; Matthias Zwicker
COMMENTS:       Code and models available at https://drive.google.com/file/d/1EzKRxnFG1Hd8g6Ggvt-jvKkgpDDwK2bY
HIGHLIGHT:      We introduce a new generative model based on denoising density estimators (DDEs), which are scalar functions parameterized by neural networks, that are efficiently trained to represent kernel density estimators of the data.

35, TITLE:      A Multigrid Method for Efficiently Training Video Models
http://arxiv.org/abs/1912.00998
AUTHORS:        Chao-Yuan Wu ; Ross Girshick ; Kaiming He ; Christoph Feichtenhofer ; Philipp Krähenbühl
COMMENTS:       CVPR 2020
HIGHLIGHT:      Inspired by multigrid methods in numerical optimization, we propose to use variable mini-batch shapes with different spatial-temporal resolutions that are varied according to a schedule.

36, TITLE:      Contextual Policy Transfer in Reinforcement Learning Domains via Deep Mixtures-of-Experts
http://arxiv.org/abs/2003.00203
AUTHORS:        Michael Gimelfarb ; Scott Sanner ; Chi-Guhn Lee
COMMENTS:       - updated experiment for Lander domain (fixed a bug in the UCB baseline) - minor editing and formatting, fixing typos - new template - 15 pages, 6 figures
HIGHLIGHT:      In this paper, we assume knowledge of estimated source task dynamics and policies, and common sub-goals but different dynamics.

37, TITLE:      X-Stance: A Multilingual Multi-Target Dataset for Stance Detection
http://arxiv.org/abs/2003.08385
AUTHORS:        Jannis Vamvas ; Rico Sennrich
COMMENTS:       SwissText + KONVENS 2020. Data and code are available at https://github.com/ZurichNLP/xstance
HIGHLIGHT:      Unlike stance detection models that have specific target issues, we use the dataset to train a single model on all the issues.

38, TITLE:      From Arguments to Key Points: Towards Automatic Argument Summarization
http://arxiv.org/abs/2005.01619
AUTHORS:        Roy Bar-Haim ; Lilach Eden ; Roni Friedman ; Yoav Kantor ; Dan Lahav ; Noam Slonim
COMMENTS:       ACL 2020
HIGHLIGHT:      We propose to represent such summaries as a small set of talking points, termed "key points", each scored according to its salience. We study the task of argument-to-key point mapping, and introduce a novel large-scale dataset for this task.

39, TITLE:      Memory-Based Graph Networks
http://arxiv.org/abs/2002.09518
AUTHORS:        Amir Hosein Khasahmadi ; Kaveh Hassani ; Parsa Moradi ; Leo Lee ; Quaid Morris
COMMENTS:       ICLR 2020
HIGHLIGHT:      We introduce an efficient memory layer for GNNs that can jointly learn node representations and coarsen the graph.

40, TITLE:      Misinformation Has High Perplexity
http://arxiv.org/abs/2006.04666
AUTHORS:        Nayeon Lee ; Yejin Bang ; Andrea Madotto ; Pascale Fung
HIGHLIGHT:      In this paper, we postulate that misinformation itself has higher perplexity compared to truthful statements, and propose to leverage the perplexity to debunk false claims in an unsupervised manner. We construct two new COVID-19-related test sets, one is scientific, and another is political in content, and empirically verify that our system performs favorably compared to existing systems. We are releasing these datasets publicly to encourage more research in debunking misinformation on COVID-19 and other topics.

41, TITLE:      Path-Sensitive Atomic Commit: Local Coordination Avoidance for Distributed Transactions
http://arxiv.org/abs/1908.05940
AUTHORS:        Tim Soethout ; Tijs van der Storm ; Jurgen Vinju
HIGHLIGHT:      Approach: In this paper we introduce Path-Sensitive Atomic Commit (PSAC) to address this situation.

42, TITLE:      Adaptive convolutional neural networks for k-space data interpolation in fast magnetic resonance imaging
http://arxiv.org/abs/2006.01385
AUTHORS:        Tianming Du ; Honggang Zhang ; Yuemeng Li ; Hee Kwon Song ; Yong Fan
HIGHLIGHT:      To overcome such limitations, we develop a deep learning algorithm, referred to as adaptive convolutional neural networks for k-space data interpolation (ACNN-k-Space), which adopts a residual Encoder-Decoder network architecture to interpolate the undersampled k-space data by integrating spatially contiguous slices as multi-channel input, along with k-space data from multiple coils if available.

43, TITLE:      Uncertainty-Aware Blind Image Quality Assessment in the Laboratory and Wild
http://arxiv.org/abs/2005.13983
AUTHORS:        Weixia Zhang ; Kede Ma ; Guangtao Zhai ; Xiaokang Yang
COMMENTS:       Under review. The implementations are available at https://github.com/zwx8981/UNIQUE
HIGHLIGHT:      To confront the cross-distortion-scenario challenge, we develop a unified BIQA model and an effective approach of training it for both synthetic and realistic distortions.

44, TITLE:      Hybrid Style Siamese Network: Incorporating style loss in complementary apparels retrieval
http://arxiv.org/abs/1912.05014
AUTHORS:        Mayukh Bhattacharyya ; Sayan Nag
COMMENTS:       Paper Accepted in the Third Workshop on Computer Vision for Fashion, Art and Design, CVPR 2020
HIGHLIGHT:      This paper proposes a mechanism of utilising those methods in this retrieval task and capturing the low level style features through a hybrid siamese network coupled with a hybrid loss.

45, TITLE:      SoftFlow: Probabilistic Framework for Normalizing Flow on Manifolds
http://arxiv.org/abs/2006.04604
AUTHORS:        Hyeongju Kim ; Hyeonseung Lee ; Woo Hyun Kang ; Joun Yeop Lee ; Nam Soo Kim
COMMENTS:       17 pages, 15figures
HIGHLIGHT:      In this paper, we propose SoftFlow, a probabilistic framework for training normalizing flows on manifolds.

46, TITLE:      An Application of Deep Reinforcement Learning to Algorithmic Trading
http://arxiv.org/abs/2004.06627
AUTHORS:        Thibaut Théate ; Damien Ernst
COMMENTS:       Preprint submitted to Elsevier journal "Expert Systems with Applications"
HIGHLIGHT:      This scientific research paper presents an innovative approach based on deep reinforcement learning (DRL) to solve the algorithmic trading problem of determining the optimal trading position at any point in time during a trading activity in stock markets.

47, TITLE:      InterBERT: An Effective Multi-Modal Pretraining Approach via Vision-and-Language Interaction
http://arxiv.org/abs/2003.13198
AUTHORS:        Junyang Lin ; An Yang ; Yichang Zhang ; Jie Liu ; Jingren Zhou ; Hongxia Yang
COMMENTS:       15 pages
HIGHLIGHT:      We propose a novel method for multi-modal pretraining, namely InterBERT (BERT for Interaction).

48, TITLE:      Conservative Agency via Attainable Utility Preservation
http://arxiv.org/abs/1902.09725
AUTHORS:        Alexander Matt Turner ; Dylan Hadfield-Menell ; Prasad Tadepalli
COMMENTS:       Published in AI, Ethics, and Society 2020
HIGHLIGHT:      To mitigate this risk, we introduce an approach that balances optimization of the primary reward function with preservation of the ability to optimize auxiliary reward functions.

49, TITLE:      A Diffractive Neural Network with Weight-Noise-Injection Training
http://arxiv.org/abs/2006.04462
AUTHORS:        Jiashuo Shi
HIGHLIGHT:      We propose a diffractive neural network with strong robustness based on Weight Noise Injection training, which achieves accurate and fast optical-based classification while diffraction layers have a certain amount of surface shape error.

50, TITLE:      Frequency-Tuned Universal Adversarial Attacks
http://arxiv.org/abs/2003.05549
AUTHORS:        Yingpeng Deng ; Lina J. Karam
HIGHLIGHT:      Based on this, we propose a frequency-tuned universal attack method to compute universal perturbations and show that our method can realize a good balance between perceivability and effectiveness in terms of fooling rate by adapting the perturbations to the local frequency content.

51, TITLE:      A Review of Emergency Incident Prediction, Resource Allocation and Dispatch Models
http://arxiv.org/abs/2006.04200
AUTHORS:        Ayan Mukhopadhyay ; Geoffrey Pettet ; Sayyed Vazirizade ; Yevgeniy Vorobeychik ; Mykel Kochenderfer ; Abhishek Dubey
HIGHLIGHT:      In this survey, we present models for incident prediction, resource allocation and dispatch concerning urban emergency incidents like accidents and crimes.