2020.06.18.txt

==========New Papers==========
1, TITLE:       DCAF: A Dynamic Computation Allocation Framework for Online Serving System
http://arxiv.org/abs/2006.09684
AUTHORS:        Biye Jiang ; Pengye Zhang ; Rihan Chen ; Binding Dai ; Xinchen Luo ; Yin Yang ; Guan Wang ; Guorui Zhou ; Xiaoqiang Zhu ; Kun Gai
HIGHLIGHT:      In this paper, we introduce a novel idea that online serving system could treat each traffic request differently and allocate "personalized" computation resource based on its value.

2, TITLE:       On the Complexity of Solving Generic Over-determined Bilinear Systems
http://arxiv.org/abs/2006.09442
AUTHORS:        John B. Baena ; Daniel Cabarcas ; Javier Verbel
HIGHLIGHT:      In this paper, we study the complexity of solving generic over-determined bilinear systems over a finite field $\mathbb{F}$.

3, TITLE:       Exploiting Review Neighbors for Contextualized Helpfulness Prediction
http://arxiv.org/abs/2006.09685
AUTHORS:        Jiahua Du ; Jia Rong ; Hua Wang ; Yanchun Zhang
HIGHLIGHT:      This paper proposes a new methodology to capture the missing interaction between reviews and their neighbors.

4, TITLE:       The Role of Verb Semantics in Hungarian Verb-Object Order
http://arxiv.org/abs/2006.09432
AUTHORS:        Dorottya Demszky ; László Kálmán ; Dan Jurafsky ; Beth Levin
COMMENTS:       Work in progress
HIGHLIGHT:      In order to investigate the role of lexical semantics in determining Hungarian word order, we conduct a large-scale, data-driven analysis on the ordering of 380 transitive verbs and their objects, as observed in hundreds of thousands of examples extracted from the Hungarian Gigaword Corpus.

5, TITLE:       Revealing the Invisible with Model and Data Shrinking for Composite-database Micro-expression Recognition
http://arxiv.org/abs/2006.09674
AUTHORS:        Zhaoqiang Xia ; Wei Peng ; Huai-Qian Khor ; Xiaoyi Feng ; Guoying Zhao
HIGHLIGHT:      In this paper, we analyze the influence of learning complexity, including the input complexity and model complexity, and discover that the lower-resolution input data and shallower-architecture model are helpful to ease the degradation of deep models in composite-database task.

6, TITLE:       A Real-time Action Representation with Temporal Encoding and Deep Compression
http://arxiv.org/abs/2006.09675
AUTHORS:        Kun Liu ; Wu Liu ; Huadong Ma ; Mingkui Tan ; Chuang Gan
HIGHLIGHT:      To address this challenge, we propose a new real-time convolutional architecture, called Temporal Convolutional 3D Network (T-C3D), for action representation.

7, TITLE:       StatAssist & GradBoost: A Study on Optimal INT8 Quantization-aware Training from Scratch
http://arxiv.org/abs/2006.09679
AUTHORS:        Taehoon Kim ; Youngjoon Yoo ; Jihoon Yang
HIGHLIGHT:      Here, we propose critical but straightforward optimization methods which enable the scratch training: floating-point statistic assisting (StatAssist) and stochastic-gradient boosting (GradBoost).

8, TITLE:       LazyIter: A Fast Algorithm for Counting Markov Equivalent DAGs and Designing Experiments
http://arxiv.org/abs/2006.09670
AUTHORS:        Ali AhmadiTeshnizi ; Saber Salehkaleybar ; Negar Kiyavash
COMMENTS:       11 pages, 2 figures, ICML
HIGHLIGHT:      We propose a method for efficient iteration over possible MECs given intervention results.

9, TITLE:       Response by the Montreal AI Ethics Institute to the European Commission's Whitepaper on AI
http://arxiv.org/abs/2006.09428
AUTHORS:        Abhishek Gupta ; Camylle Lanteigne
COMMENTS:       Submitted to the European Commission
HIGHLIGHT:      This paper outlines the EC's policy options for the promotion and adoption of artificial intelligence (AI) in the European Union.

10, TITLE:      MetaSDF: Meta-learning Signed Distance Functions
http://arxiv.org/abs/2006.09662
AUTHORS:        Vincent Sitzmann ; Eric R. Chan ; Richard Tucker ; Noah Snavely ; Gordon Wetzstein
COMMENTS:       Project website: https://vsitzmann.github.io/metasdf/
HIGHLIGHT:      Here, we formalize learning of a shape space as a meta-learning problem and leverage gradient-based meta-learning algorithms to solve this task.

11, TITLE:      Implicit Neural Representations with Periodic Activation Functions
http://arxiv.org/abs/2006.09661
AUTHORS:        Vincent Sitzmann ; Julien N. P. Martel ; Alexander W. Bergman ; David B. Lindell ; Gordon Wetzstein
COMMENTS:       Project website: https://vsitzmann.github.io/siren/ Project video: https://youtu.be/Q2fLWGBeaiI
HIGHLIGHT:      We propose to leverage periodic activation functions for implicit neural representations and demonstrate that these networks, dubbed sinusoidal representation networks or Sirens, are ideally suited for representing complex natural signals and their derivatives.

12, TITLE:      Ranking and benchmarking framework for sampling algorithms on synthetic data streams
http://arxiv.org/abs/2006.09895
AUTHORS:        József Dániel Gáspár ; Martin Horváth ; Győző Horváth ; Zoltán Zvara
HIGHLIGHT:      To tackle this imbalance, we can use dynamic partitioning algorithms that require a sampling algorithm to precisely estimate the underlying distribution of the data stream.

13, TITLE:      On the Learnability of Concepts: With Applications to Comparing Word Embedding Algorithms
http://arxiv.org/abs/2006.09896
AUTHORS:        Adam Sutton ; Nello Cristianini
COMMENTS:       7 Pages. AIAI 2020. 5 equations 6 tables
HIGHLIGHT:      In this paper we introduce the notion of "concept" as a list of words that have shared semantic content.

14, TITLE:      Self-Supervised Representation Learning for Visual Anomaly Detection
http://arxiv.org/abs/2006.09654
AUTHORS:        Rabia Ali ; Muhammad Umar Karim Khan ; Chong Min Kyung
HIGHLIGHT:      We consider the problem of anomaly detection in images and videos, and present a new visual anomaly detection technique for videos.

15, TITLE:      Fairness-Oriented Semi-Chaotic Genetic Algorithm-Based Channel Assignment Technique for Nodes Starvation Problem in Wireless Mesh Network
http://arxiv.org/abs/2006.09655
AUTHORS:        Fuad A. Ghaleb ; Bander Ali Saleh Al-rimy ; Maznah Kamat ; Mohd. Foad Rohani ; Shukor Abd Razak
COMMENTS:       18 pages, 10 Figures
HIGHLIGHT:      To this end, the Fairness-Oriented Semi-Chaotic Genetic Algorithm-Based Channel Assignment Technique (FA-SCGA-CAA) was proposed in this paper for Nodes Starvation Problem in Wireless Mesh Networks.

16, TITLE:      Diverse Rule Sets
http://arxiv.org/abs/2006.09890
AUTHORS:        Guangyi Zhang ; Aristides Gionis
HIGHLIGHT:      Here we propose a novel approach of inferring diverse rule sets, by optimizing small overlap among decision rules with a 2-approximation guarantee under the framework of Max-Sum diversification.

17, TITLE:      Fine-grained Sentiment Controlled Text Generation
http://arxiv.org/abs/2006.09891
AUTHORS:        Bidisha Samanta ; Mohit Agarwal ; Niloy Ganguly
HIGHLIGHT:      Controlled text generation techniques aim to regulate specific attributes (e.g. sentiment) while preserving the attribute independent content.

18, TITLE:      Unsupervised Learning of Visual Features by Contrasting Cluster Assignments
http://arxiv.org/abs/2006.09882
AUTHORS:        Mathilde Caron ; Ishan Misra ; Julien Mairal ; Priya Goyal ; Piotr Bojanowski ; Armand Joulin
HIGHLIGHT:      In this paper, we propose an online algorithm, SwAV, that takes advantage of contrastive methods without requiring to compute pairwise comparisons.

19, TITLE:      Automatic Curriculum Learning through Value Disagreement
http://arxiv.org/abs/2006.09641
AUTHORS:        Yunzhi Zhang ; Pieter Abbeel ; Lerrel Pinto
COMMENTS:       https://sites.google.com/berkeley.edu/vds/
HIGHLIGHT:      Inspired by this, we propose setting up an automatic curriculum for goals that the agent needs to solve.

20, TITLE:      Parameterized MDPs and Reinforcement Learning Problems -- A Maximum Entropy Principle Based Framework
http://arxiv.org/abs/2006.09646
AUTHORS:        Amber Srivastava ; Srinivasa M Salapaka
COMMENTS:       13 pages, 4 figures
HIGHLIGHT:      We present a framework to address a class of sequential decision making problems.

21, TITLE:      Using Wavelets and Spectral Methods to Study Patterns in Image-Classification Datasets
http://arxiv.org/abs/2006.09879
AUTHORS:        Roozbeh Yousefzadeh ; Furong Huang
HIGHLIGHT:      In this paper, we use wavelet transformation and spectral methods to analyze the contents of image classification datasets, extract specific patterns from the datasets and find the associations between patterns and classes.

22, TITLE:      Iterative Edit-Based Unsupervised Sentence Simplification
http://arxiv.org/abs/2006.09639
AUTHORS:        Dhruv Kumar ; Lili Mou ; Lukasz Golab ; Olga Vechtomova
COMMENTS:       The paper has been accepted to ACL 2020
HIGHLIGHT:      We present a novel iterative, edit-based approach to unsupervised sentence simplification.

23, TITLE:      An Exploratory Study of Argumentative Writing by Young Students: A Transformer-based Approach
http://arxiv.org/abs/2006.09873
AUTHORS:        Debanjan Ghosh ; Beata Beigman Klebanov ; Yi Song
COMMENTS:       15th Workshop on Innovative Use of NLP for Building Educational Applications, ACL 2020
HIGHLIGHT:      We present a computational exploration of argument critique writing by young students.

24, TITLE:      Deeply Learned Spectral Total Variation Decomposition
http://arxiv.org/abs/2006.10004
AUTHORS:        Tamara G. Grossmann ; Yury Korolev ; Guy Gilboa ; Carola-Bibiane Schönlieb
HIGHLIGHT:      In this paper, we present a neural network approximation of a non-linear spectral decomposition.

25, TITLE:      Self-Supervised Joint Learning Framework of Depth Estimation via Implicit Cues
http://arxiv.org/abs/2006.09876
AUTHORS:        Jianrong Wang ; Ge Zhang ; Zhenyu Wu ; XueWei Li ; Li Liu
HIGHLIGHT:      In this work, we propose a novel self-supervised joint learning framework for depth estimation using consecutive frames from monocular and stereo videos.

26, TITLE:      Logic, Probability and Action: A Situation Calculus Perspective
http://arxiv.org/abs/2006.09868
AUTHORS:        Vaishak Belle
HIGHLIGHT:      In this paper, we survey recent results pertaining to the integration of logic, probability and actions in the situation calculus, which is arguably one of the oldest and most well-known formalisms.

27, TITLE:      Building Low-Resource NER Models Using Non-Speaker Annotation
http://arxiv.org/abs/2006.09627
AUTHORS:        Tatiana Tsygankova ; Francesca Marini ; Stephen Mayhew ; Dan Roth
HIGHLIGHT:      In this work we propose an alternative approach to building low-resource Named Entity Recognition (NER) models using "non-speaker" (NS) annotations, provided by annotators with no prior experience in the target language.

28, TITLE:      Visor: Privacy-Preserving Video Analytics as a Cloud Service
http://arxiv.org/abs/2006.09628
AUTHORS:        Rishabh Poddar ; Ganesh Ananthanarayanan ; Srinath Setty ; Stavros Volos ; Raluca Ada Popa
COMMENTS:       USENIX Security 2020
HIGHLIGHT:      We present Visor, a system that provides confidentiality for the user's video stream as well as the ML models in the presence of a compromised cloud platform and untrusted co-tenants.

29, TITLE:      Fast Object Classification and Meaningful Data Representation of Segmented Lidar Instances
http://arxiv.org/abs/2006.10011
AUTHORS:        Lukas Hahn ; Frederik Hasecke ; Anton Kummert
COMMENTS:       6 pages, 5 figures, 4 tables; accepted to appear in IEEE ITSC 2020
HIGHLIGHT:      In this work, we propose a way to facilitate real-time Lidar object classification on CPU.

30, TITLE:      Learning Visual Commonsense for Robust Scene Graph Generation
http://arxiv.org/abs/2006.09623
AUTHORS:        Alireza Zareian ; Haoxuan You ; Zhecan Wang ; Shih-Fu Chang
HIGHLIGHT:      We propose the first method to acquire visual commonsense such as affordance and intuitive physics automatically from data, and use that to enhance scene graph generation.

31, TITLE:      Intelligent Protection & Classification of Transients in Two-Core Symmetric Phase Angle Regulating Transformers
http://arxiv.org/abs/2006.09865
AUTHORS:        Pallav Kumar Bera ; Can Isik
HIGHLIGHT:      This paper investigates the applicability of time and time-frequency features based classifiers to distinguish internal faults and other transients - magnetizing inrush, sympathetic inrush, external faults with current transformer saturation, and overexcitation - for Indirect Symmetrical Phase Angle Regulating Transformers (ISPAR).

32, TITLE:      De-Anonymizing Text by Fingerprinting Language Generation
http://arxiv.org/abs/2006.09615
AUTHORS:        Zhen Sun ; Roei Schuster ; Vitaly Shmatikov
HIGHLIGHT:      We initiate the study of code security of ML systems by investigating how nucleus sampling---a popular approach for generating text, used for applications such as auto-completion---unwittingly leaks texts typed by users.

33, TITLE:      Dynamic Tensor Rematerialization
http://arxiv.org/abs/2006.09616
AUTHORS:        Marisa Kirisame ; Steven Lyubomirsky ; Altan Haan ; Jennifer Brennan ; Mike He ; Jared Roesch ; Tianqi Chen ; Zachary Tatlock
COMMENTS:       28 pages, 11 figures, implementation available here: https://github.com/uwsampl/dtr-prototype
HIGHLIGHT:      We present Dynamic Tensor Rematerialization (DTR), a greedy online algorithm for heuristically checkpointing arbitrary models.

34, TITLE:      Multi-Subspace Neural Network for Image Recognition
http://arxiv.org/abs/2006.09618
AUTHORS:        Chieh-Ning Fang ; Chin-Teng Lin
HIGHLIGHT:      In this study, we proposed multi-subspace neural network (MSNN) which integrates key components of the convolutional neural network (CNN), receptive field, with subspace concept.

35, TITLE:      Generalising Recursive Neural Models by Tensor Decomposition
http://arxiv.org/abs/2006.10021
AUTHORS:        Daniele Castellana ; Davide Bacciu
COMMENTS:       Accepted at IEEE WCCI2020
HIGHLIGHT:      In this work we introduce a general approach to model aggregation of structural context leveraging a tensor-based formulation.

36, TITLE:      Conversational Neuro-Symbolic Commonsense Reasoning
http://arxiv.org/abs/2006.10022
AUTHORS:        Forough Arabshahi ; Jennifer Lee ; Mikayla Gawarecki ; Kathryn Mazaitis ; Amos Azaria ; Tom Mitchell
HIGHLIGHT:      We propose a new commonsense reasoning benchmark where the task is to uncover commonsense presumptions implied by imprecisely stated natural language commands in the form of if-then-because statements. We release a benchmark data set for this task, collected from humans and annotated with commonsense presumptions.

37, TITLE:      Canonicalizing Open Knowledge Bases with Multi-Layered Meta-Graph Neural Network
http://arxiv.org/abs/2006.09610
AUTHORS:        Tianwen Jiang ; Tong Zhao ; Bing Qin ; Ting Liu ; Nitesh V. Chawla ; Meng Jiang
HIGHLIGHT:      In this work, we integrate structural information (from which tuple, which sentence) and semantic information (semantic similarity) to do the canonicalization.

38, TITLE:      Shallow Feature Based Dense Attention Network for Crowd Counting
http://arxiv.org/abs/2006.09853
AUTHORS:        Yunqi Miao ; Zijia Lin ; Guiguang Ding ; Jungong Han
HIGHLIGHT:      In this paper, we propose a Shallow feature based Dense Attention Network (SDANet) for crowd counting from still images, which diminishes the impact of backgrounds via involving a shallow feature based attention model, and meanwhile, captures multi-scale information via densely connecting hierarchical image features.

39, TITLE:      Big Self-Supervised Models are Strong Semi-Supervised Learners
http://arxiv.org/abs/2006.10029
AUTHORS:        Ting Chen ; Simon Kornblith ; Kevin Swersky ; Mohammad Norouzi ; Geoffrey Hinton
COMMENTS:       code and pretrained models at https://github.com/google-research/simclr
HIGHLIGHT:      We find that, the fewer the labels, the more this approach (task-agnostic use of unlabeled data) benefits from a bigger network.

40, TITLE:      Landscape-Aware Fixed-Budget Performance Regression and Algorithm Selection for Modular CMA-ES Variants
http://arxiv.org/abs/2006.09855
AUTHORS:        Anja Jankovic ; Carola Doerr
COMMENTS:       To appear in Proc. of Genetic and Evolutionary Computation Conference (GECCO'20)
HIGHLIGHT:      We test this approach on a very challenging problem: algorithm selection on a portfolio of very similar algorithms, which we choose from the family of modular CMA-ES algorithms.

41, TITLE:      Task-agnostic Exploration in Reinforcement Learning
http://arxiv.org/abs/2006.09497
AUTHORS:        Xuezhou Zhang ; Yuzhe ma ; Adish Singla
HIGHLIGHT:      We present an efficient task-agnostic RL algorithm, \textsc{UCBZero}, that finds $\epsilon$-optimal policies for $N$ arbitrary tasks after at most $\tilde O(\log(N)H^5SA/\epsilon^2)$ exploration episodes.

42, TITLE:      EPIE Dataset: A Corpus For Possible Idiomatic Expressions
http://arxiv.org/abs/2006.09479
AUTHORS:        Prateek Saxena ; Soma Paul
HIGHLIGHT:      With this in mind, we present our English Possible Idiomatic Expressions(EPIE) corpus containing 25206 sentences labelled with lexical instances of 717 idiomatic expressions.

43, TITLE:      Selective Question Answering under Domain Shift
http://arxiv.org/abs/2006.09462
AUTHORS:        Amita Kamath ; Robin Jia ; Percy Liang
COMMENTS:       ACL 2020
HIGHLIGHT:      In this work, we propose the setting of selective question answering under domain shift, in which a QA model is tested on a mixture of in-domain and out-of-domain data, and must answer (i.e., not abstain on) as many questions as possible while maintaining high accuracy.

44, TITLE:      Interpretable multimodal fusion networks reveal mechanisms of brain cognition
http://arxiv.org/abs/2006.09454
AUTHORS:        Wenxing Hu ; Xianghe Meng ; Yuntong Bai ; Aiying Zhang ; Biao Cai ; Gemeng Zhang ; Tony W. Wilson ; Julia M. Stephen ; Vince D. Calhoun ; Yu-Ping Wang
HIGHLIGHT:      In this work, we develop an interpretable multimodal fusion model, namely gCAM-CCL, which can perform automated diagnosis and result interpretation simultaneously.

45, TITLE:      Noise2Inpaint: Learning Referenceless Denoising by Inpainting Unrolling
http://arxiv.org/abs/2006.09450
AUTHORS:        Burhaneddin Yaman ; Seyed Amir Hossein Hosseini ; Mehmet Akçakaya
HIGHLIGHT:      In this work, building on these approaches and recent methods from image reconstruction, we introduce Noise2Inpaint (N2I), a training approach that recasts the denoising problem into a regularized image inpainting framework.

46, TITLE:      3D Shape Reconstruction from Free-Hand Sketches
http://arxiv.org/abs/2006.09694
AUTHORS:        Jiayun Wang ; Jierui Lin ; Qian Yu ; Runtao Liu ; Yubei Chen ; Stella X. Yu
HIGHLIGHT:      We pioneer to study this task and aim to enhance the power of sketches in 3D-related applications such as interactive design and VR/AR games.

47, TITLE:      Simplified Swarm Optimization for Bi-Objection Active Reliability Redundancy Allocation Problems
http://arxiv.org/abs/2006.09844
AUTHORS:        Wei-Chang Yeh
HIGHLIGHT:      In this study, a bi-objective RRAP is formulated by changing the cost constraint as a new goal, because it is necessary to balance the reliability and cost impact for the entire system in practical applications.

48, TITLE:      Burst Photography for Learning to Enhance Extremely Dark Images
http://arxiv.org/abs/2006.09845
AUTHORS:        Ahmet Serdar Karadeniz ; Erkut Erdem ; Aykut Erdem
HIGHLIGHT:      Motivated by these studies, in this paper, we aim to leverage burst photography to boost the performance and obtain much sharper and more accurate RGB images from extremely dark raw images.

49, TITLE:      Learning Sparse Masks for Efficient Image Super-Resolution
http://arxiv.org/abs/2006.09603
AUTHORS:        Longguang Wang ; Xiaoyu Dong ; Yingqian Wang ; Xinyi Ying ; Zaiping Lin ; Wei An ; Yulan Guo
HIGHLIGHT:      To address this limitation, we develop an SR network (SMSR) to learn sparse masks to prune redundant computation conditioned on the input image.

50, TITLE:      Multi-Domain Level Generation and Blending with Sketches via Example-Driven BSP and Variational Autoencoders
http://arxiv.org/abs/2006.09807
AUTHORS:        Sam Snodgrass ; Anurag Sarkar
COMMENTS:       To appear in FDG 2020 Cite as: @inproceedings{snodgrass2020blending, title={Multi-Domain Level Generation and Blending with Sketches via Example-Driven BSP and Variational Autoencoders}, author={Snodgrass, Sam and Sarkar, Anurag}, booktitle={Proceedings of the 15th International Conference on the Foundations of Digital Games}, year={2020} }
HIGHLIGHT:      In this paper, we present a PCGML approach for level generation that is able to recombine, adapt, and reuse structural patterns from several domains to approximate unseen domains.

51, TITLE:      Mining Label Distribution Drift in Unsupervised Domain Adaptation
http://arxiv.org/abs/2006.09565
AUTHORS:        Peizhao Li ; Zhengming Ding ; Hongfu Liu
HIGHLIGHT:      Numerical results and empirical model analysis show that LMDAN delivers superior performance compared to other state-of-the-art domain adaptation methods under such scenarios.

52, TITLE:      Explanation-based Weakly-supervised Learning of Visual Relations with Graph Networks
http://arxiv.org/abs/2006.09562
AUTHORS:        Federico Baldassarre ; Kevin Smith ; Josephine Sullivan ; Hossein Azizpour
HIGHLIGHT:      This paper introduces a novel weakly-supervised method for visual relationship detection that relies only on image-level predicate annotations.

53, TITLE:      Optimizing Grouped Convolutions on Edge Devices
http://arxiv.org/abs/2006.09791
AUTHORS:        Perry Gibson ; José Cano ; Jack Turner ; Elliot J. Crowley ; Michael O'Boyle ; Amos Storkey
COMMENTS:       Camera ready version to be published at ASAP 2020 - The 31st IEEE International Conference on Application-specific Systems, Architectures and Processors. 8 pages, 6 figures
HIGHLIGHT:      In this paper we propose Grouped Spatial Pack Convolutions (GSPC), a new implementation of grouped convolutions that outperforms existing solutions.

54, TITLE:      Learning to Learn with Feedback and Local Plasticity
http://arxiv.org/abs/2006.09549
AUTHORS:        Jack Lindsey ; Ashok Litwin-Kumar
HIGHLIGHT:      In this study, we employ meta-learning to discover networks that learn using feedback connections and local, biologically inspired learning rules.

55, TITLE:      Self-supervised Knowledge Distillation for Few-shot Learning
http://arxiv.org/abs/2006.09785
AUTHORS:        Jathushan Rajasegaran ; Salman Khan ; Munawar Hayat ; Fahad Shahbaz Khan ; Mubarak Shah
HIGHLIGHT:      In this paper, we propose a simple approach to improve the representation capacity of deep neural networks for few-shot learning tasks.

56, TITLE:      Reinforcement Learning with Uncertainty Estimation for Tactical Decision-Making in Intersections
http://arxiv.org/abs/2006.09786
AUTHORS:        Carl-Johan Hoel ; Tommy Tram ; Jonas Sjöberg
HIGHLIGHT:      In this study, the uncertainty information is used to choose safe actions in unknown situations, which removes all collisions from within the training distribution, and most collisions outside of the distribution.

57, TITLE:      Sketch-Guided Scenery Image Outpainting
http://arxiv.org/abs/2006.09788
AUTHORS:        Yaxiong Wang ; Yunchao Wei ; Xueming Qian ; Li Zhu ; Yi Yang
COMMENTS:       conditional image outpainting
HIGHLIGHT:      In this work, we take the image outpainting one step forward by allowing users to harvest personal custom outpainting results using sketches as the guidance.

58, TITLE:      COLREG-Compliant Collision Avoidance for Unmanned Surface Vehicle using Deep Reinforcement Learning
http://arxiv.org/abs/2006.09540
AUTHORS:        Eivind Meyer ; Amalie Heiberg ; Adil Rasheed ; Omer San
HIGHLIGHT:      For many decades, they have been subject to academic study, leading to a vast number of proposed approaches.

59, TITLE:      Mitosis Detection Under Limited Annotation: A Joint Learning Approach
http://arxiv.org/abs/2006.09772
AUTHORS:        Pushpak Pati ; Antonio Foncubierta-Rodriguez ; Orcun Goksel ; Maria Gabrani
COMMENTS:       2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI)
HIGHLIGHT:      We propose a deep classification framework for enhancing mitosis detection by leveraging class label information, via softmax loss, and spatial distribution information among samples, via distance metric learning.

60, TITLE:      Quality Management of Machine Learning Systems
http://arxiv.org/abs/2006.09529
AUTHORS:        P. Santhanam
COMMENTS:       AAAI-20 Workshop on Engineering Dependable and Secure Machine Learning Systems- February 7, 2020
HIGHLIGHT:      The purpose of this paper is to present a view of a holistic quality management framework for ML applications based on the current advances and identify new areas of software engineering research to achieve a more trustworthy AI.

61, TITLE:      Evaluation of 3D CNN Semantic Mapping for Rover Navigation
http://arxiv.org/abs/2006.09761
AUTHORS:        Sebastiano Chiodini ; Luca Torresin ; Marco Pertile ; Stefano Debei
COMMENTS:       To be presented at the 7th IEEE International Workshop on Metrology for Aerospace (MetroAerospace)
HIGHLIGHT:      In this work we present a technique to generate accurate three-dimensional semantic maps for Martian environment.

62, TITLE:      Maximum Roaming Multi-Task Learning
http://arxiv.org/abs/2006.09762
AUTHORS:        Lucas Pascal ; Pietro Michiardi ; Xavier Bost ; Benoit Huet ; Maria A. Zuluaga
HIGHLIGHT:      In this work, we present a novel way to partition the parameter space without weakening the inductive bias.

63, TITLE:      Improving unsupervised neural aspect extraction for online discussions using out-of-domain classification
http://arxiv.org/abs/2006.09766
AUTHORS:        Anton Alekseev ; Elena Tutubalina ; Valentin Malykh ; Sergey Nikolenko
COMMENTS:       Journal of Intelligent & Fuzzy Systems, pre-press, https://content.iospress.com/articles/journal-of-intelligent-and-fuzzy-systems/ifs179908
HIGHLIGHT:      In this work, we introduce a simple approach based on sentence filtering in order to improve topical aspects learned from newsgroups-based content without modifying the basic mechanism of ABAE.

64, TITLE:      Cross-lingual Retrieval for Iterative Self-Supervised Training
http://arxiv.org/abs/2006.09526
AUTHORS:        Chau Tran ; Yuqing Tang ; Xian Li ; Jiatao Gu
HIGHLIGHT:      In this work, we found that the cross-lingual alignment can be further improved by training seq2seq models on sentence pairs mined using their own encoder outputs.

65, TITLE:      Aligning with Heterogeneous Preferences for Kidney Exchange
http://arxiv.org/abs/2006.09519
AUTHORS:        Rachel Freedman
COMMENTS:       Presented at the IJCAI-PRICAI 2020 Workshop on Artificial Intelligence Safety
HIGHLIGHT:      In this paper, we address this problem in a real-world public health context: kidney exchange.

66, TITLE:      Noise or Signal: The Role of Image Backgrounds in Object Recognition
http://arxiv.org/abs/2006.09994
AUTHORS:        Kai Xiao ; Logan Engstrom ; Andrew Ilyas ; Aleksander Madry
HIGHLIGHT:      We create a toolkit for disentangling foreground and background signal on ImageNet images, and find that (a) models can achieve non-trivial accuracy by relying on the background alone, (b) models often misclassify images even in the presence of correctly classified foregrounds--up to 87.5% of the time with adversarially chosen backgrounds, and (c) more accurate models tend to depend on backgrounds less.

67, TITLE:      On sparse connectivity, adversarial robustness, and a novel model of the artificial neuron
http://arxiv.org/abs/2006.09510
AUTHORS:        Sergey Bochkanov
HIGHLIGHT:      In this paper, we propose two closely connected methods to improve these metrics on contour recognition tasks: (a) a novel model of an artificial neuron, a "strong neuron," with low hardware requirements and inherent robustness against adversarial perturbations and (b) a novel constructive training algorithm that generates sparse networks with $O(1)$ connections per neuron.

68, TITLE:      Visual Chirality
http://arxiv.org/abs/2006.09512
AUTHORS:        Zhiqiu Lin ; Jin Sun ; Abe Davis ; Noah Snavely
COMMENTS:       Published at CVPR 2020, Best Paper Nomination, Oral Presentation. Project Page: https://linzhiqiu.github.io/papers/chirality/
HIGHLIGHT:      In this paper, we investigate how the statistics of visual data are changed by reflection.

69, TITLE:      Universal Lower-Bounds on Classification Error under Adversarial Attacks and Random Corruption
http://arxiv.org/abs/2006.09989
AUTHORS:        Elvis Dohmatob
HIGHLIGHT:      Our contributions are three-fold.

70, TITLE:      Solving the Order Batching and Sequencing Problem using Deep Reinforcement Learning
http://arxiv.org/abs/2006.09507
AUTHORS:        Bram Cals ; Yingqian Zhang ; Remco Dijkman ; Claudy van Dorst
COMMENTS:       Preprint
HIGHLIGHT:      In this paper, we present a Deep Reinforcement Learning (DRL) approach for deciding how and when orders should be batched and picked in a warehouse to minimize the number of tardy orders.

71, TITLE:      Probabilistic orientation estimation with matrix Fisher distributions
http://arxiv.org/abs/2006.09740
AUTHORS:        D. Mohlin ; G. Bianchi ; J. Sullivan
COMMENTS:       20 pages, 11 figures, submitted to NeurIPS
HIGHLIGHT:      We overcome this issue by using a neural network to output the parameters for a matrix Fisher distribution since these parameters are homeomorphic to $\mathbb{R}^9$.

72, TITLE:      On the Inference of Soft Biometrics from Typing Patterns Collected in a Multi-device Environment
http://arxiv.org/abs/2006.09501
AUTHORS:        Vishaal Udandarao ; Mohit Agrawal ; Rajesh Kumar ; Rajiv Ratn Shah
COMMENTS:       The first two authors contributed equally. The code is available upon request. Please contact the last author
HIGHLIGHT:      In this paper, we study the inference of gender, major/minor (computer science, non-computer science), typing style, age, and height from the typing patterns collected from 117 individuals in a multi-device environment.

73, TITLE:      A generalizable saliency map-based interpretation of model outcome
http://arxiv.org/abs/2006.09504
AUTHORS:        Shailja Thakur ; Sebastian Fischmeister
HIGHLIGHT:      To fully exploit the capabilities of complex neural networks, we propose a non-intrusive interpretability technique that uses the input and output of the model to generate a saliency map.

74, TITLE:      LRPD: Long Range 3D Pedestrian Detection Leveraging Specific Strengths of LiDAR and RGB
http://arxiv.org/abs/2006.09738
AUTHORS:        Michael Fürst ; Oliver Wasenmüller ; Didier Stricker
COMMENTS:       7 Pages, 5 Figures, Autonomous Vehicles, 3D Object Detection
HIGHLIGHT:      Thus, we propose an approach specifically targeting long range 3D pedestrian detection (LRPD), leveraging the density of RGB and the precision of LiDAR.

75, TITLE:      Breaking Type-Safety in Go: An Empirical Study on the Usage of the unsafe Package
http://arxiv.org/abs/2006.09973
AUTHORS:        Diego Elias Costa ; Suhaib Mujahid ; Rabe Abdalkareem ; Emad Shihab
HIGHLIGHT:      In this paper, we present the first large-scale study on the usage of the unsafe package in 2,438 popular Go projects.

76, TITLE:      Playing Unique Games on Certified Small-Set Expanders
http://arxiv.org/abs/2006.09969
AUTHORS:        Mitali Bafna ; Boaz Barak ; Pravesh Kothari ; Tselil Schramm ; David Steurer
HIGHLIGHT:      We give an algorithm for solving unique games (UG) instances whose constraints correspond to edges of graphs with a sum-of-squares (SoS) small-set-expansion certificate.

77, TITLE:      Cross-Correlated Attention Networks for Person Re-Identification
http://arxiv.org/abs/2006.09597
AUTHORS:        Jieming Zhou ; Soumava Kumar Roy ; Pengfei Fang ; Mehrtash Harandi ; Lars Petersson
COMMENTS:       Accepted by Image and Vision Computing
HIGHLIGHT:      In this paper, we propose a new attention module called Cross-Correlated Attention (CCA); which aims to overcome such limitations by maximizing the information gain between different attended regions.

78, TITLE:      CO-Search: COVID-19 Information Retrieval with Semantic Search, Question Answering, and Abstractive Summarization
http://arxiv.org/abs/2006.09595
AUTHORS:        Andre Esteva ; Anuprit Kale ; Romain Paulus ; Kazuma Hashimoto ; Wenpeng Yin ; Dragomir Radev ; Richard Socher
HIGHLIGHT:      Here we present CO-Search, a retriever-ranker semantic search engine designed to handle complex queries over the COVID-19 literature, potentially aiding overburdened health workers in finding scientific answers during a time of crisis.

79, TITLE:      Modeling subjective assessments of guilt in newspaper crime narratives
http://arxiv.org/abs/2006.09589
AUTHORS:        Elisa Kreiss ; Zijian Wang ; Christopher Potts
HIGHLIGHT:      Such models might be used as tools for understanding the societal effects of crime reporting.

80, TITLE:      Semantic Visual Navigation by Watching YouTube Videos
http://arxiv.org/abs/2006.10034
AUTHORS:        Matthew Chang ; Arjun Gupta ; Saurabh Gupta
COMMENTS:       Project website with videos: https://matthewchang.github.io/value-learning-from-videos/
HIGHLIGHT:      We improve upon end-to-end RL methods by 66%, while using 250x fewer interactions.

81, TITLE:      WhoAmI: An Automatic Tool for Visual Recognition of Tiger and Leopard Individuals in the Wild
http://arxiv.org/abs/2006.09962
AUTHORS:        Rita Pucci ; Jitendra Shankaraiah ; Devcharan Jathanna ; Ullas Karanth ; Kartic Subr
HIGHLIGHT:      we propose the first fully-automatic tool that can recognize specific individuals of leopard and tiger due to their characteristic body markings.

82, TITLE:      LSD-C: Linearly Separable Deep Clusters
http://arxiv.org/abs/2006.10039
AUTHORS:        Sylvestre-Alvise Rebuffi ; Sebastien Ehrhardt ; Kai Han ; Andrea Vedaldi ; Andrew Zisserman
COMMENTS:       Code available at https://github.com/srebuffi/lsd-clusters
HIGHLIGHT:      We present LSD-C, a novel method to identify clusters in an unlabeled dataset.

83, TITLE:      A Tweet-based Dataset for Company-Level Stock Return Prediction
http://arxiv.org/abs/2006.09723
AUTHORS:        Karolina Sowinska ; Pranava Madhyastha
COMMENTS:       Dataset available here: https://github.com/ImperialNLP/stockreturnpred
HIGHLIGHT:      In this paper, we present a dataset that allows for company-level analysis of tweet based impact on one-, two-, three-, and seven-day stock returns. Our dataset consists of 862, 231 labelled instances from twitter in English, we also release a cleaned subset of 85, 176 labelled instances to the community.

84, TITLE:      High-Fidelity Generative Image Compression
http://arxiv.org/abs/2006.09965
AUTHORS:        Fabian Mentzer ; George Toderici ; Michael Tschannen ; Eirikur Agustsson
COMMENTS:       Project page: https://hific.github.io
HIGHLIGHT:      We extensively study how to combine Generative Adversarial Networks and learned compression to obtain a state-of-the-art generative lossy compression system.

85, TITLE:      When We First Met: Visual-Inertial Person Localization for Co-Robot Rendezvous
http://arxiv.org/abs/2006.09959
AUTHORS:        Xi Sun ; Xinshuo Weng ; Kris Kitani
HIGHLIGHT:      We propose a method to learn a visual-inertial feature space in which the motion of a person in video can be easily matched to the motion measured by a wearable inertial measurement unit (IMU).

86, TITLE:      Neural Anisotropy Directions
http://arxiv.org/abs/2006.09717
AUTHORS:        Guillermo Ortiz-Jimenez ; Apostolos Modas ; Seyed-Mohsen Moosavi-Dezfooli ; Pascal Frossard
COMMENTS:       39 pages, 22 figures
HIGHLIGHT:      In this work, we analyze the role of the network architecture in shaping the inductive bias of deep classifiers.

87, TITLE:      Learning to Detect 3D Reflection Symmetry for Single-View Reconstruction
http://arxiv.org/abs/2006.10042
AUTHORS:        Yichao Zhou ; Shichen Liu ; Yi Ma
HIGHLIGHT:      In this work, we focus on object-level 3D reconstruction and present a geometry-based end-to-end deep learning framework that first detects the mirror plane of reflection symmetry that commonly exists in man-made objects and then predicts depth maps by finding the intra-image pixel-wise correspondence of the symmetry.

88, TITLE:      Automatically Ranked Russian Paraphrase Corpus for Text Generation
http://arxiv.org/abs/2006.09719
AUTHORS:        Vadim Gudkov ; Olga Mitrofanova ; Elizaveta Filippskikh
COMMENTS:       To be published in The 4th Workshop on Neural Generation and Translation @ ACL 2020
HIGHLIGHT:      The article is focused on automatic development and ranking of a large corpus for Russian paraphrase generation which proves to be the first corpus of such type in Russian computational linguistics.

89, TITLE:      Delta Schema Network in Model-based Reinforcement Learning
http://arxiv.org/abs/2006.09950
AUTHORS:        Andrey Gorodetskiy ; Alexandra Shlychkova ; Aleksandr I. Panov
HIGHLIGHT:      In the paper we are expanding the schema networks method which allows to extract the logical relationships between objects and actions from the environment data.

90, TITLE:      Universally Quantized Neural Compression
http://arxiv.org/abs/2006.09952
AUTHORS:        Eirikur Agustsson ; Lucas Theis
COMMENTS:       Authors contributed equally
HIGHLIGHT:      We demonstrate that a uniform noise channel can also be implemented at test time using universal quantization (Ziv, 1985).

91, TITLE:      Adversarial Defense by Latent Style Transformations
http://arxiv.org/abs/2006.09701
AUTHORS:        Shuo Wang ; Surya Nepal ; Marthie Grobler ; Carsten Rudolph ; Tianle Chen ; Shangyu Chen
HIGHLIGHT:      In this paper, we investigate an attack-agnostic defense against adversarial attacks on high-resolution images by detecting suspicious inputs. We then build a set of edited copies with non-essential style transformations by performing latent shifting and reconstruction, based on the correspondences between latent codes and style transformations.

92, TITLE:      Forgetful Experience Replay in Hierarchical Reinforcement Learning from Demonstrations
http://arxiv.org/abs/2006.09939
AUTHORS:        Alexey Skrynnik ; Aleksey Staroverov ; Ermek Aitygulov ; Kirill Aksenov ; Vasilii Davydov ; Aleksandr I. Panov
HIGHLIGHT:      In this paper, we propose a combination of these approaches that allow the agent to use low-quality demonstrations in complex vision-based environments with multiple related goals.

93, TITLE:      CoSE: Compositional Stroke Embeddings
http://arxiv.org/abs/2006.09930
AUTHORS:        Emre Aksan ; Thomas Deselaers ; Andrea Tagliasacchi ; Otmar Hilliges
HIGHLIGHT:      We present a generative model for stroke-based drawing tasks which is able to model complex free-form structures.

94, TITLE:      Contrastive Learning for Weakly Supervised Phrase Grounding
http://arxiv.org/abs/2006.09920
AUTHORS:        Tanmay Gupta ; Arash Vahdat ; Gal Chechik ; Xiaodong Yang ; Jan Kautz ; Derek Hoiem
HIGHLIGHT:      We show that phrase grounding can be learned by optimizing word-region attention to maximize a lower bound on mutual information between images and caption words.

95, TITLE:      FISHING Net: Future Inference of Semantic Heatmaps In Grids
http://arxiv.org/abs/2006.09917
AUTHORS:        Noureldin Hendy ; Cooper Sloan ; Feng Tian ; Pengfei Duan ; Nick Charchut ; Yuesong Xie ; Chuang Wang ; James Philbin
HIGHLIGHT:      In this work, we present an end-to-end pipeline that performs semantic segmentation and short term prediction using a top-down representation.

96, TITLE:      Green Simulation Assisted Reinforcement Learning with Model Risk for Biomanufacturing Learning and Control
http://arxiv.org/abs/2006.09919
AUTHORS:        Hua Zheng ; Wei Xie ; Mingbin Ben Feng
COMMENTS:       12 pages, 1 figures. To appear in the Proceedings of the 2020 Winter Simulation Conference (WSC)
HIGHLIGHT:      To address these challenges, we propose a green simulation assisted model-based reinforcement learning to support process online learning and guide dynamic decision making.

97, TITLE:      Vision-Aided Dynamic Blockage Prediction for 6G Wireless Communication Networks
http://arxiv.org/abs/2006.09902
AUTHORS:        Gouranga Charan ; Muhammad Alrabeiah ; Ahmed Alkhateeb
COMMENTS:       The dataset and code files will be available soon on the ViWi website: https://www.viwi-dataset.net/
HIGHLIGHT:      It proposes a novel solution that proactively predicts \textit{dynamic} link blockages.

98, TITLE:      Learning Colour Representations of Search Queries
http://arxiv.org/abs/2006.09904
AUTHORS:        Paridhi Maheshwari ; Manoj Ghuhan ; Vishwa Vinay
COMMENTS:       Accepted as a full paper at SIGIR 2020
HIGHLIGHT:      In this work, we consider the role of colour in this relevance matching process.



==========Updates to Previous Papers==========
1, TITLE:       How to Train Your Super-Net: An Analysis of Training Heuristics in Weight-Sharing NAS
http://arxiv.org/abs/2003.04276
AUTHORS:        Kaicheng Yu ; Rene Ranftl ; Mathieu Salzmann
COMMENTS:       Updated with latest results on NASBench-101, now we achieve 0.48 sparse Kendall-Tau on this space
HIGHLIGHT:      In this paper, we therefore provide a systematic evaluation of the heuristics and hyperparameters that are frequently employed by weight-sharing NAS algorithms.

2, TITLE:       PointVoteNet: Accurate Object Detection and 6 DoF Pose Estimation in Point Clouds
http://arxiv.org/abs/1912.09057
AUTHORS:        Frederik Hagelskjær ; Anders Glent Buch
COMMENTS:       5 pages
HIGHLIGHT:      We present a learning-based method for 6 DoF pose estimation of rigid objects in point cloud data.

3, TITLE:       Equilibrium Propagation for Complete Directed Neural Networks
http://arxiv.org/abs/2006.08798
AUTHORS:        Matilde Tristany Farinha ; Sérgio Pequito ; Pedro A. Santos ; Mário A. T. Figueiredo
COMMENTS:       6 pages, 6 images, accepted for ESANN 2020
HIGHLIGHT:      Specifically, we introduce: a new neuronal dynamics and learning rule for arbitrary network architectures; a sparsity-inducing method able to prune irrelevant connections; a dynamical-systems characterization of the models, using Lyapunov theory.

4, TITLE:       Explainable Active Learning (XAL): An Empirical Study of How Local Explanations Impact Annotator Experience
http://arxiv.org/abs/2001.09219
AUTHORS:        Bhavya Ghai ; Q. Vera Liao ; Yunfeng Zhang ; Rachel Bellamy ; Klaus Mueller
COMMENTS:       working draft. replacing the first draft with a correction on workload data and additional analysis on individual differences
HIGHLIGHT:      Toward this vision, we propose a novel paradigm of explainable active learning (XAL), by introducing techniques from the recently surging field of explainable AI (XAI) into an AL setting.

5, TITLE:       Towards reusable network components by learning compatible representations
http://arxiv.org/abs/2004.03898
AUTHORS:        Michael Gygli ; Jasper Uijlings ; Vittorio Ferrari
HIGHLIGHT:      In particular, we split a network into two components, a features extractor and a target task head, and propose various approaches to accomplish compatibility between them.

6, TITLE:       Consistency of circuit lower bounds with bounded theories
http://arxiv.org/abs/1905.12935
AUTHORS:        Jan Bydzovsky ; Jan Krajicek ; Igor C. Oliveira
HIGHLIGHT:      Motivated by this and related questions about the interaction between mathematical proofs and computations, we investigate circuit complexity from the perspective of logic.

7, TITLE:       Machine Learning Classification Informed by a Functional Biophysical System
http://arxiv.org/abs/1911.08589
AUTHORS:        Jason A. Platt ; Anna Miller ; Lawson Fuller ; Henry D. I. Abarbanel
HIGHLIGHT:      We present a novel machine learning architecture for classification suggested by experiments on olfactory systems.

8, TITLE:       A Spatio-temporal Transformer for 3D Human Motion Prediction
http://arxiv.org/abs/2004.08692
AUTHORS:        Emre Aksan ; Peng Cao ; Manuel Kaufmann ; Otmar Hilliges
COMMENTS:       New baselines in the main result table
HIGHLIGHT:      In this paper, we propose a novel Transformer-based architecture for the task of generative modelling of 3D human motion.

9, TITLE:       Featherweight Go
http://arxiv.org/abs/2005.11710
AUTHORS:        Robert Griesemer ; Raymond Hu ; Wen Kokke ; Julien Lange ; Ian Lance Taylor ; Bernardo Toninho ; Philip Wadler ; Nobuko Yoshida
COMMENTS:       Full version
HIGHLIGHT:      We describe a design for generics in Go inspired by previous work on Featherweight Java by Igarashi, Pierce, and Wadler.

10, TITLE:      Neural gradients are lognormally distributed: understanding sparse and quantized training
http://arxiv.org/abs/2006.08173
AUTHORS:        Brian Chmiel ; Liad Ben-Uri ; Moran Shkolnik ; Elad Hoffer ; Ron Banner ; Daniel Soudry
COMMENTS:       Fix references typos
HIGHLIGHT:      Taking this into account, we suggest two methods to reduce the computational and memory burdens of neural gradients.

11, TITLE:      SpineNet: Learning Scale-Permuted Backbone for Recognition and Localization
http://arxiv.org/abs/1912.05027
AUTHORS:        Xianzhi Du ; Tsung-Yi Lin ; Pengchong Jin ; Golnaz Ghiasi ; Mingxing Tan ; Yin Cui ; Quoc V. Le ; Xiaodan Song
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we argue encoder-decoder architecture is ineffective in generating strong multi-scale features because of the scale-decreased backbone.

12, TITLE:      Symbiosis Promotes Fitness Improvements in the Game of Life
http://arxiv.org/abs/1908.07034
AUTHORS:        Peter D. Turney
COMMENTS:       Changes to Sections 1, 3, 4, 5, and 6. Figures and tables appear at the end of the document
HIGHLIGHT:      We present a computational simulation of evolving entities that includes symbiosis with shifting levels of selection.

13, TITLE:      Learning Query Inseparable ELH Ontologies
http://arxiv.org/abs/1911.07229
AUTHORS:        Ana Ozaki ; Cosimo Persia ; Andrea Mazzullo
HIGHLIGHT:      We investigate the complexity of learning query inseparable ELH ontologies in a variant of Angluin's exact learning model.

14, TITLE:      SEEK: Segmented Embedding of Knowledge Graphs
http://arxiv.org/abs/2005.00856
AUTHORS:        Wentao Xu ; Shun Zheng ; Liang He ; Bin Shao ; Jian Yin ; Tie-Yan Liu
HIGHLIGHT:      To mitigate this problem, we propose a lightweight modeling framework that can achieve highly competitive relational expressiveness without increasing the model complexity.

15, TITLE:      Robust registration of medical images in the presence of spatially-varying noise
http://arxiv.org/abs/1711.04247
AUTHORS:        Reza Abbasi-Asl ; Aboozar Ghaffari ; Emad Fatemizadeh
HIGHLIGHT:      We show that the spatially-varying noise is highly expressed in the residual component of the EMD and could be filtered out.

16, TITLE:      Automated diagnosis of COVID-19 with limited posteroanterior chest X-ray images using fine-tuned deep neural networks
http://arxiv.org/abs/2004.11676
AUTHORS:        Narinder Singh Punn ; Sonali Agarwal
HIGHLIGHT:      Following from this context, this article presents the random oversampling and weighted class loss function approach for unbiased fine-tuned learning (transfer learning) in various state-of-the-art deep learning approaches such as baseline ResNet, Inception-v3, Inception ResNet-v2, DenseNet169, and NASNetLarge to perform binary classification (as normal and COVID-19 cases) and also multi-class classification (as COVID-19, pneumonia, and normal case) of posteroanterior CXR images.

17, TITLE:      A Hybrid Natural Language Generation System Integrating Rules and Deep Learning Algorithms
http://arxiv.org/abs/2006.09213
AUTHORS:        Wei Wei ; Bei Zhou ; Georgios Leontidis
COMMENTS:       6 pages
HIGHLIGHT:      This paper proposes an enhanced natural language generation system combining the merits of both rule-based approaches and modern deep learning algorithms, boosting its performance to the extent where the generated textual content is capable of exhibiting agile human-writing styles and the content logic of which is highly controllable.

18, TITLE:      A Collaborative Ecosystem for Digital Coptic Studies
http://arxiv.org/abs/1912.05082
AUTHORS:        Caroline T. Schroeder ; Amir Zeldes
COMMENTS:       9 pages; paper presented at the Stanford University CESTA Workshop "Collecting, Preserving and Disseminating Endangered Cultural Heritage for New Understandings Through Multilingual Approaches"
HIGHLIGHT:      In this paper, we outline some of the latest developments in Coptic Scriptorium, a DH project dedicated to bringing Coptic resources online in uniform, machine readable, and openly available formats.

19, TITLE:      Tamil Vowel Recognition With Augmented MNIST-like Data Set
http://arxiv.org/abs/2006.08367
AUTHORS:        Muthiah Annamalai
COMMENTS:       8 pages, 3 figures, 4 tables
HIGHLIGHT:      We report generation of a MNIST [4] compatible data set [1] for Tamil vowels to enable building a classification DNN or other such ML/AI deep learning [2] models for Tamil OCR/Handwriting applications.

20, TITLE:      MobileGAN: Skin Lesion Segmentation Using a Lightweight Generative Adversarial Network
http://arxiv.org/abs/1907.00856
AUTHORS:        Md. Mostafa Kamal Sarker ; Hatem A. Rashwan ; Mohamed Abdel-Nasser ; Vivek Kumar Singh ; Syeda Furruka Banu ; Farhan Akram ; Forhad U H Chowdhury ; Kabir Ahmed Choudhury ; Sylvie Chambon ; Petia Radeva ; Domenec Puig
COMMENTS:       30 pages, Submitted to Expert Systems with Applications
HIGHLIGHT:      Thus, this paper aims at achieving high precise segmentation with minimum resources by presenting a lightweight and efficient generative adversarial network (GAN) model, called MobileGAN.

21, TITLE:      A Smooth Representation of Belief over SO(3) for Deep Rotation Learning with Uncertainty
http://arxiv.org/abs/2006.01031
AUTHORS:        Valentin Peretroukhin ; Matthew Giamou ; David M. Rosen ; W. Nicholas Greene ; Nicholas Roy ; Jonathan Kelly
COMMENTS:       Accepted to Robotics: Science and Systems (RSS'20), Boston, Massachusetts, USA, Jul. 12-16, 2020
HIGHLIGHT:      In this work, we present a novel symmetric matrix representation of the 3D rotation group, SO(3), with two important properties that make it particularly suitable for learned models: (1) it satisfies a smoothness property that improves convergence and generalization when regressing large rotation targets, and (2) it encodes a symmetric Bingham belief over the space of unit quaternions, permitting the training of uncertainty-aware models.

22, TITLE:      BIAS: Transparent reporting of biomedical image analysis challenges
http://arxiv.org/abs/1910.04071
AUTHORS:        Lena Maier-Hein ; Annika Reinke ; Michal Kozubek ; Anne L. Martel ; Tal Arbel ; Matthias Eisenmann ; Allan Hanbuary ; Pierre Jannin ; Henning Müller ; Sinan Onogur ; Julio Saez-Rodriguez ; Bram van Ginneken ; Annette Kopp-Schneider ; Bennett Landman
COMMENTS:       2 Appendices - Appendix A: BIAS reporting guideline for biomedical image analysis challenges, Appendix B: Glossary 2 Supplements - Suppl 1: Form for summarizing information on challenge organization, Suppl 2: Structured description of a challenge design
HIGHLIGHT:      This article describes how the BIAS statement was developed and presents a checklist which authors of biomedical image analysis challenges are encouraged to include in their submission when giving a paper on a challenge into review.

23, TITLE:      A Characteristic Function Approach to Deep Implicit Generative Modeling
http://arxiv.org/abs/1909.07425
AUTHORS:        Abdul Fatir Ansari ; Jonathan Scarlett ; Harold Soh
COMMENTS:       CVPR 2020 (Oral), Code available at https://github.com/clear-nus/OCFGAN
HIGHLIGHT:      In this paper, we formulate the problem of learning an IGM as minimizing the expected distance between characteristic functions.

24, TITLE:      Robust Quantization: One Model to Rule Them All
http://arxiv.org/abs/2002.07686
AUTHORS:        Moran Shkolnik ; Brian Chmiel ; Ron Banner ; Gil Shomron ; Yury Nahshan ; Alex Bronstein ; Uri Weiser
HIGHLIGHT:      To address this issue, we propose a method that provides intrinsic robustness to the model against a broad range of quantization processes.

25, TITLE:      Deep Learning on Image Denoising: An overview
http://arxiv.org/abs/1912.13171
AUTHORS:        Chunwei Tian ; Lunke Fei ; Wenxian Zheng ; Yong Xu ; Wangmeng Zuo ; Chia-Wen Lin
HIGHLIGHT:      In this paper, we offer a comparative study of deep techniques in image denoising.

26, TITLE:      microPhantom: Playing microRTS under uncertainty and chaos
http://arxiv.org/abs/2005.11019
AUTHORS:        Florian Richoux
HIGHLIGHT:      In this paper, we focus on decision-making under uncertainty, by tackling the Unit Production Problem with a method based on a combination of Constraint Programming and decision theory.

27, TITLE:      Stable Backward Diffusion Models that Minimise Convex Energies
http://arxiv.org/abs/1903.03491
AUTHORS:        Leif Bergerhoff ; Marcelo Cárdenas ; Joachim Weickert ; Martin Welk
HIGHLIGHT:      It is therefore greatly desirable to establish a backward diffusion model which implements a smart stabilisation approach that can be used in combination with an easy to handle numerical scheme.

28, TITLE:      $R^3$: Reverse, Retrieve, and Rank for Sarcasm Generation with Commonsense Knowledge
http://arxiv.org/abs/2004.13248
AUTHORS:        Tuhin Chakrabarty ; Debanjan Ghosh ; Smaranda Muresan ; Nanyun Peng
COMMENTS:       Accepted at the 2020 Annual Conference of the Association for Computational Linguistics (ACL)
HIGHLIGHT:      We propose an unsupervised approach for sarcasm generation based on a non-sarcastic input sentence.

29, TITLE:      VATEX: A Large-Scale, High-Quality Multilingual Dataset for Video-and-Language Research
http://arxiv.org/abs/1904.03493
AUTHORS:        Xin Wang ; Jiawei Wu ; Junkun Chen ; Lei Li ; Yuan-Fang Wang ; William Yang Wang
COMMENTS:       ICCV 2019 Oral. 17 pages, 14 figures, 6 tables (updated the VATEX website link: vatex-challenge.org)
HIGHLIGHT:      We present a new large-scale multilingual video description dataset, VATEX, which contains over 41,250 videos and 825,000 captions in both English and Chinese.

30, TITLE:      Causality Extraction based on Self-Attentive BiLSTM-CRF with Transferred Embeddings
http://arxiv.org/abs/1904.07629
AUTHORS:        Zhaoning Li ; Qi Li ; Xiaotian Zou ; Jiangtao Ren
COMMENTS:       39 pages, 11 figures, 6 tables
HIGHLIGHT:      In this paper, we formulate causality extraction as a sequence labeling problem based on a novel causality tagging scheme.

31, TITLE:      StarNet: towards weakly supervised few-shot detection and explainable few-shot classification
http://arxiv.org/abs/2003.06798
AUTHORS:        Leonid Karlinsky ; Joseph Shtok ; Amit Alfassy ; Moshe Lichtenstein ; Sivan Harary ; Eli Schwartz ; Sivan Doveh ; Prasanna Sattigeri ; Rogerio Feris ; Alexander Bronstein ; Raja Giryes
HIGHLIGHT:      In this paper, we introduce StarNet, featuring an end-to-end differentiable non-parametric star-model classification head.

32, TITLE:      Marginal Utility for Planning in Continuous or Large Discrete Action Spaces
http://arxiv.org/abs/2006.06054
AUTHORS:        Zaheen Farraz Ahmad ; Levi H. S. Lelis ; Michael Bowling
HIGHLIGHT:      In this paper we explore explicitly learning a candidate action generator by optimizing a novel objective, marginal utility.

33, TITLE:      Smooth Contextual Bandits: Bridging the Parametric and Non-differentiable Regret Regimes
http://arxiv.org/abs/1909.02553
AUTHORS:        Yichun Hu ; Nathan Kallus ; Xiaojie Mao
HIGHLIGHT:      We study a nonparametric contextual bandit problem where the expected reward functions belong to a H\"older class with smoothness parameter $\beta$.

34, TITLE:      Feature Space Saturation during Training
http://arxiv.org/abs/2006.08679
AUTHORS:        Justin Shenk ; Mats L. Richter ; Wolf Byttner ; Anders Arpteg ; Mikael Huss
COMMENTS:       23 pages, 26 figures, fix citation formatting, add link highlighting
HIGHLIGHT:      We propose a computationally lightweight method for approximating the variance matrix during training.

35, TITLE:      Implicit Generation and Generalization in Energy-Based Models
http://arxiv.org/abs/1903.08689
AUTHORS:        Yilun Du ; Igor Mordatch
HIGHLIGHT:      We present techniques to scale MCMC based EBM training on continuous neural networks, and we show its success on the high-dimensional data domains of ImageNet32x32, ImageNet128x128, CIFAR-10, and robotic hand trajectories, achieving better samples than other likelihood models and nearing the performance of contemporary GAN approaches, while covering all modes of the data.

36, TITLE:      On Sharpness of Error Bounds for Multivariate Neural Network Approximation
http://arxiv.org/abs/2004.02203
AUTHORS:        Steffen Goebbels
COMMENTS:       ongoing work
HIGHLIGHT:      The paper deals with best non-linear approximation by such sums of ridge functions.

37, TITLE:      Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?
http://arxiv.org/abs/2003.11539
AUTHORS:        Yonglong Tian ; Yue Wang ; Dilip Krishnan ; Joshua B. Tenenbaum ; Phillip Isola
COMMENTS:       First two authors contributed equally. Project Page: https://people.csail.mit.edu/yuewang/projects/rfs/ Code: http://github.com/WangYueFt/rfs/
HIGHLIGHT:      In this work, we show that a simple baseline: learning a supervised or self-supervised representation on the meta-training set, followed by training a linear classifier on top of this representation, outperforms state-of-the-art few-shot learning methods.

38, TITLE:      Exposing Backdoors in Robust Machine Learning Models
http://arxiv.org/abs/2003.00865
AUTHORS:        Ezekiel Soremekun ; Sakshi Udeshi ; Sudipta Chattopadhyay
HIGHLIGHT:      In this paper, we demonstrate that adversarially robust models are susceptible to backdoor attacks.

39, TITLE:      Complexity Analysis of an Edge Preserving CNN SAR Despeckling Algorithm
http://arxiv.org/abs/2004.08345
AUTHORS:        Sergio Vitale ; Giampaolo Ferraioli ; Vito Pascazio
COMMENTS:       Accpeted to International Geoscience and Remote Sensing Symposium, IGARSS 2020
HIGHLIGHT:      In the last decades several methods for SAR denoising have been proposed and in the last years great attention has moved towards deep learning based solutions.

40, TITLE:      Constraints in Developing a Complete Bengali Optical Character Recognition System
http://arxiv.org/abs/2003.08384
AUTHORS:        Abu Saleh Md. Abir ; Sanjana Rahman ; Samia Ellin ; Maisha Farzana ; Md Hridoy Manik ; Chowdhury Rafeed Rahman
HIGHLIGHT:      Constraints in Developing a Complete Bengali Optical Character Recognition System

41, TITLE:      DeepCoDA: personalized interpretability for compositional health data
http://arxiv.org/abs/2006.01392
AUTHORS:        Thomas P. Quinn ; Dang Nguyen ; Santu Rana ; Sunil Gupta ; Svetha Venkatesh
COMMENTS:       To appear at ICML 2020
HIGHLIGHT:      We propose the Deep Compositional Data Analysis (DeepCoDA) framework to extend precision health modelling to high-dimensional compositional data, and to provide personalized interpretability through patient-specific weights.

42, TITLE:      Distill, Adapt, Distill: Training Small, In-Domain Models for Neural Machine Translation
http://arxiv.org/abs/2003.02877
AUTHORS:        Mitchell A. Gordon ; Kevin Duh
COMMENTS:       Accepted to WNGT 2020 Workshop at ACL 2020 Conference
HIGHLIGHT:      We explore best practices for training small, memory efficient machine translation models with sequence-level knowledge distillation in the domain adaptation setting.

43, TITLE:      Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere
http://arxiv.org/abs/2005.10242
AUTHORS:        Tongzhou Wang ; Phillip Isola
COMMENTS:       International Conference on Machine Learning (ICML), 2020
HIGHLIGHT:      In this work, we identify two key properties related to the contrastive loss: (1) alignment (closeness) of features from positive pairs, and (2) uniformity of the induced distribution of the (normalized) features on the hypersphere.

44, TITLE:      Effective and Efficient Dropout for Deep Convolutional Neural Networks
http://arxiv.org/abs/1904.03392
AUTHORS:        Shaofeng Cai ; Yao Shu ; Gang Chen ; Beng Chin Ooi ; Wei Wang ; Meihui Zhang
COMMENTS:       12 pages, 10 figures
HIGHLIGHT:      In this paper, we revisit this issue and examine various dropout variants in an attempt to improve existing dropout-based regularization techniques for CNNs.

45, TITLE:      ktrain: A Low-Code Library for Augmented Machine Learning
http://arxiv.org/abs/2004.10703
AUTHORS:        Arun S. Maiya
COMMENTS:       9 pages
HIGHLIGHT:      We present ktrain, a low-code Python library that makes machine learning more accessible and easier to apply.

46, TITLE:      Self-explaining AI as an alternative to interpretable AI
http://arxiv.org/abs/2002.05149
AUTHORS:        Daniel C. Elton
COMMENTS:       16pgs. Shorter version to appear in Proceedings of the 13th Annual Conference on Artificial General Intelligence (AGI-2020)
HIGHLIGHT:      To show how we might be able to trust AI despite these problems we introduce the concept of self-explaining AI.

47, TITLE:      Maximum Probability Theorem: A Framework for Probabilistic Learning
http://arxiv.org/abs/1910.09417
AUTHORS:        Amir Emad Marvasti ; Ehsan Emad Marvasti ; Hassan Foroosh
HIGHLIGHT:      We present a theoretical framework of probabilistic learning derived by Maximum Probability (MP) Theorem shown in the current paper.

48, TITLE:      3D Dynamic Scene Graphs: Actionable Spatial Perception with Places, Objects, and Humans
http://arxiv.org/abs/2002.06289
AUTHORS:        Antoni Rosinol ; Arjun Gupta ; Marcus Abate ; Jingnan Shi ; Luca Carlone
COMMENTS:       11 pages, 5 figures
HIGHLIGHT:      We present a unified representation for actionable spatial perception: 3D Dynamic Scene Graphs.

49, TITLE:      Controllable Person Image Synthesis with Attribute-Decomposed GAN
http://arxiv.org/abs/2003.12267
AUTHORS:        Yifang Men ; Yiming Mao ; Yuning Jiang ; Wei-Ying Ma ; Zhouhui Lian
COMMENTS:       Accepted by CVPR 2020 (Oral). Project Page: https://menyifang.github.io/projects/ADGAN/ADGAN.html
HIGHLIGHT:      This paper introduces the Attribute-Decomposed GAN, a novel generative model for controllable person image synthesis, which can produce realistic person images with desired human attributes (e.g., pose, head, upper clothes and pants) provided in various source inputs.

50, TITLE:      Where is the Information in a Deep Neural Network?
http://arxiv.org/abs/1905.12213
AUTHORS:        Alessandro Achille ; Giovanni Paolini ; Stefano Soatto
HIGHLIGHT:      We establish a novel relation between the information in the weights and the effective information in the activations, and use this result to show that models with low (information) complexity not only generalize better, but are bound to learn invariant representations of future inputs.

51, TITLE:      Pattern Denoising in Molecular Associative Memory using Pairwise Markov Random Field Models
http://arxiv.org/abs/2005.13780
AUTHORS:        Dharani Punithan ; Byoung-Tak Zhang
HIGHLIGHT:      We propose an in silico molecular associative memory model for pattern learning, storage and denoising using Pairwise Markov Random Field (PMRF) model.

52, TITLE:      Mucko: Multi-Layer Cross-Modal Knowledge Reasoning for Fact-based Visual Question Answering
http://arxiv.org/abs/2006.09073
AUTHORS:        Zihao Zhu ; Jing Yu ; Yujing Wang ; Yajing Sun ; Yue Hu ; Qi Wu
COMMENTS:       Accepted by IJCAI 2020. SOLE copyright holder is IJCAI (international Joint Conferences on Artificial Intelligence)
HIGHLIGHT:      In this paper, we depict an image by a multi-modal heterogeneous graph, which contains multiple layers of information corresponding to the visual, semantic and factual features.

53, TITLE:      Gradient Descent Finds Global Minima for Generalizable Deep Neural Networks of Practical Sizes
http://arxiv.org/abs/1908.02419
AUTHORS:        Kenji Kawaguchi ; Jiaoyang Huang
COMMENTS:       Accepted. All the results remain the same. Additional explanations were added
HIGHLIGHT:      In this paper, we theoretically prove that gradient descent can find a global minimum of non-convex optimization of all layers for nonlinear deep neural networks of sizes commonly encountered in practice.

54, TITLE:      Heidelberg Colorectal Data Set for Surgical Data Science in the Sensor Operating Room
http://arxiv.org/abs/2005.03501
AUTHORS:        Lena Maier-Hein ; Martin Wagner ; Tobias Ross ; Annika Reinke ; Sebastian Bodenstedt ; Peter M. Full ; Hellena Hempe ; Diana Mindroc-Filimon ; Patrick Scholz ; Thuy Nuong Tran ; Pierangela Bruno ; Anna Kisilenko ; Benjamin Müller ; Tornike Davitashvili ; Manuela Capek ; Minu Tizabi ; Matthias Eisenmann ; Tim J. Adler ; Janek Gröhl ; Melanie Schellenberg ; Silvia Seidlitz ; T. Y. Emmy Lai ; Veith Roethlingshoefer ; Fabian Both ; Sebastian Bittel ; Marc Mengler ; Lars Mündermann ; Martin Apitz ; Stefanie Speidel ; Hannes G. Kenngott ; Beat P. Müller-Stich
COMMENTS:       Submitted to Nature Scientific Data
HIGHLIGHT:      This paper introduces the Heidelberg Colorectal (HeiCo) data set - the first publicly available data set enabling comprehensive benchmarking of medical instrument detection and segmentation algorithms with a specific emphasis on robustness and generalization capabilities of the methods.

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

56, TITLE:      The Source-Target Domain Mismatch Problem in Machine Translation
http://arxiv.org/abs/1909.13151
AUTHORS:        Jiajun Shen ; Peng-Jen Chen ; Matt Le ; Junxian He ; Jiatao Gu ; Myle Ott ; Michael Auli ; Marc'Aurelio Ranzato
HIGHLIGHT:      In this work we study the effect of local context in machine translation and postulate that particularly in low resource settings this causes the domains of the source and target language to greatly mismatch, as the two languages are often spoken in further apart regions of the world with more distinctive cultural traits and unrelated local events.

57, TITLE:      Better Set Representations For Relational Reasoning
http://arxiv.org/abs/2003.04448
AUTHORS:        Qian Huang ; Horace He ; Abhay Singh ; Yan Zhang ; Ser-Nam Lim ; Austin Benson
COMMENTS:       Preprint, 17 pages
HIGHLIGHT:      To resolve this limitation, we propose a simple and general network module called a Set Refiner Network (SRN).

58, TITLE:      BSP-Net: Generating Compact Meshes via Binary Space Partitioning
http://arxiv.org/abs/1911.06971
AUTHORS:        Zhiqin Chen ; Andrea Tagliasacchi ; Hao Zhang
COMMENTS:       CVPR 2020 Best Student Paper Award. Project page: https://bsp-net.github.io, Code: https://github.com/czq142857/BSP-NET-original
HIGHLIGHT:      Leading methods for learning generative models of shapes rely on implicit functions, and generate meshes only after expensive iso-surfacing routines.