2020.06.19.txt

==========New Papers==========
1, TITLE:       AMALGUM -- A Free, Balanced, Multilayer English Web Corpus
http://arxiv.org/abs/2006.10677
AUTHORS:        Luke Gessler ; Siyao Peng ; Yang Liu ; Yilun Zhu ; Shabnam Behzad ; Amir Zeldes
COMMENTS:       Accepted at LREC 2020. See https://www.aclweb.org/anthology/2020.lrec-1.648/ (note: ACL Anthology's title is currently out of date)
HIGHLIGHT:      We present a freely available, genre-balanced English web corpus totaling 4M tokens and featuring a large number of high-quality automatic annotation layers, including dependency trees, non-named entity annotations, coreference resolution, and discourse trees in Rhetorical Structure Theory.

2, TITLE:       Dissecting Deep Networks into an Ensemble of Generative Classifiers for Robust Predictions
http://arxiv.org/abs/2006.10679
AUTHORS:        Lokender Tiwari ; Anish Madan ; Saket Anand ; Subhashis Banerjee
COMMENTS:       Demo code available at https://github.com/lokender/REGroup
HIGHLIGHT:      In this work, our investigations of intermediate representations of a pre-trained DNN lead to an interesting discovery pointing to intrinsic robustness to adversarial attacks.

3, TITLE:       "EHLO WORLD" -- Checking If Your Conversational AI Knows Right from Wrong
http://arxiv.org/abs/2006.10437
AUTHORS:        Elayne Ruane ; Vivek Nallur
COMMENTS:       8 pages, 2 figures, SoCAI 2020 : AISB Symposium on Conversational AI
HIGHLIGHT:      In this paper we discuss approaches to evaluating and validating the ethical claims of a Conversational AI system.

4, TITLE:       BlazePose: On-device Real-time Body Pose tracking
http://arxiv.org/abs/2006.10204
AUTHORS:        Valentin Bazarevsky ; Ivan Grishchenko ; Karthik Raveendran ; Tyler Zhu ; Fan Zhang ; Matthias Grundmann
COMMENTS:       4 pages, 6 figures; CVPR Workshop on Computer Vision for Augmented and Virtual Reality, Seattle, WA, USA, 2020
HIGHLIGHT:      We present BlazePose, a lightweight convolutional neural network architecture for human pose estimation that is tailored for real-time inference on mobile devices.

5, TITLE:       Parameterized Inapproximability of Independent Set in $H$-Free Graphs
http://arxiv.org/abs/2006.10444
AUTHORS:        Pavel Dvořák ; Andreas Emil Feldmann ; Ashutosh Rai ; Paweł Rzążewski
COMMENTS:       Preliminary version of the paper in WG 2020 proceedings
HIGHLIGHT:      We study the Independent Set (IS) problem in $H$-free graphs, i.e., graphs excluding some fixed graph $H$ as an induced subgraph.

6, TITLE:       HyNet: Local Descriptor with Hybrid Similarity Measure and Triplet Loss
http://arxiv.org/abs/2006.10202
AUTHORS:        Yurun Tian ; Axel Barroso-Laguna ; Tony Ng ; Vassileios Balntas ; Krystian Mikolajczyk
HIGHLIGHT:      In this paper, we investigate how L2 normalisation affects the back-propagated descriptor gradients during training.

7, TITLE:       Political Advertising Dataset: the use case of the Polish 2020 Presidential Elections
http://arxiv.org/abs/2006.10207
AUTHORS:        Łukasz Augustyniak ; Krzysztof Rajda ; Tomasz Kajdanowicz ; Michał Bernaczyk
COMMENTS:       ACL 2020 WiNLP Workshop - accepted
HIGHLIGHT:      We present the first publicly open dataset for detecting specific text chunks and categories of political advertising in the Polish language.

8, TITLE:       Deep Multitask Learning for Pervasive BMI Estimation and Identity Recognition in Smart Beds
http://arxiv.org/abs/2006.10453
AUTHORS:        Vandad Davoodnia ; Monet Slinowsky ; Ali Etemad
COMMENTS:       This is a pre-print of an article published in journal of Ambient Intelligence and Humanized Computing. The final authenticated version is available online at: https://doi.org/10.1007/s12652-020-02210-9
HIGHLIGHT:      In this paper, simultaneous estimation and monitoring of body mass index (BMI) and user identity recognition through a unified machine learning framework using smart beds is explored.

9, TITLE:       UV-Net: Learning from Curve-Networks and Solids
http://arxiv.org/abs/2006.10211
AUTHORS:        Pradeep Kumar Jayaraman ; Aditya Sanghi ; Joseph Lambourne ; Thomas Davies ; Hooman Shayani ; Nigel Morris
HIGHLIGHT:      In this paper, we propose a unified representation for parametric curve-networks and solids by exploiting the u- and uv-parameter domains of curve and surfaces, respectively, to model the geometry, and an adjacency graph to explicitly model the topology.

10, TITLE:      Learning High-Resolution Domain-Specific Representations with a GAN Generator
http://arxiv.org/abs/2006.10451
AUTHORS:        Danil Galeev ; Konstantin Sofiiuk ; Danila Rukhovich ; Mikhail Romanov ; Olga Barinova ; Anton Konushin
HIGHLIGHT:      In this work we study representations learnt by a GAN generator.

11, TITLE:      Language Guided Networks for Cross-modal Moment Retrieval
http://arxiv.org/abs/2006.10457
AUTHORS:        Kun Liu ; Xun Yang ; Tat-seng Chua ; Huadong Ma ; Chuang Gan
HIGHLIGHT:      In this paper, we present Language Guided Networks (LGN), a new framework that tightly integrates cross-modal features in multiple stages.

12, TITLE:      Generating Fundus Fluorescence Angiography Images from Structure Fundus Images Using Generative Adversarial Networks
http://arxiv.org/abs/2006.10216
AUTHORS:        Wanyue Li ; Wen Kong ; Yiwei Chen ; Jing Wang ; Yi He ; Guohua Shi ; Guohua Deng
COMMENTS:       16 pages, 6 figures, accepted by Medical Imaging on Deep Learning
HIGHLIGHT:      In this work, we proposed a conditional generative adversarial network(GAN) - based method to directly learn the mapping relationship between structure fundus images and fundus fluorescence angiography images.

13, TITLE:      SEAL: Segment-wise Extractive-Abstractive Long-form Text Summarization
http://arxiv.org/abs/2006.10213
AUTHORS:        Yao Zhao ; Mohammad Saleh ; Peter J. Liu
HIGHLIGHT:      In this paper, we study long-form abstractive text summarization, a sequence-to-sequence setting with input sequence lengths up to 100,000 tokens and output sequence lengths up to 768 tokens.

14, TITLE:      MediaPipe Hands: On-device Real-time Hand Tracking
http://arxiv.org/abs/2006.10214
AUTHORS:        Fan Zhang ; Valentin Bazarevsky ; Andrey Vakunov ; Andrei Tkachenka ; George Sung ; Chuo-Ling Chang ; Matthias Grundmann
COMMENTS:       5 pages, 7 figures; CVPR Workshop on Computer Vision for Augmented and Virtual Reality, Seattle, WA, USA, 2020
HIGHLIGHT:      We present a real-time on-device hand tracking pipeline that predicts hand skeleton from single RGB camera for AR/VR applications.

15, TITLE:      Sequential Graph Convolutional Network for Active Learning
http://arxiv.org/abs/2006.10219
AUTHORS:        Razvan Caramalau ; Binod Bhattarai ; Tae-Kyun Kim
HIGHLIGHT:      We propose a novel generic sequential Graph Convolution Network (GCN) training for Active Learning.

16, TITLE:      STEAM: Self-Supervised Taxonomy Expansion with Mini-Paths
http://arxiv.org/abs/2006.10217
AUTHORS:        Yue Yu ; Yinghao Li ; Jiaming Shen ; Hao Feng ; Jimeng Sun ; Chao Zhang
COMMENTS:       KDD 2020 Research Track Full Paper
HIGHLIGHT:      We propose a self-supervised taxonomy expansion model named STEAM, which leverages natural supervision in the existing taxonomy for expansion.

17, TITLE:      SXL: Spatially explicit learning of geographic processes with auxiliary tasks
http://arxiv.org/abs/2006.10461
AUTHORS:        Konstantin Klemmer ; Daniel B. Neill
HIGHLIGHT:      We introduce SXL, a method for learning with geospatial data using explicitly spatial auxiliary tasks.

18, TITLE:      Efficient Execution of Quantized Deep Learning Models: A Compiler Approach
http://arxiv.org/abs/2006.10226
AUTHORS:        Animesh Jain ; Shoubhik Bhattacharya ; Masahiro Masuda ; Vin Sharma ; Yida Wang
HIGHLIGHT:      In this paper, we address the challenges of executing quantized deep learning models on diverse hardware platforms by proposing an augmented compiler approach.

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

20, TITLE:      Progressively Unfreezing Perceptual GAN
http://arxiv.org/abs/2006.10250
AUTHORS:        Jinxuan Sun ; Yang Chen ; Junyu Dong ; Guoqiang Zhong
HIGHLIGHT:      In this paper, we propose a general framework, called Progressively Unfreezing Perceptual GAN (PUPGAN), which can generate images with fine texture details.

21, TITLE:      Video Moment Localization using Object Evidence and Reverse Captioning
http://arxiv.org/abs/2006.10260
AUTHORS:        Madhawa Vidanapathirana ; Supriya Pandhre ; Sonia Raychaudhuri ; Anjali Khurana
COMMENTS:       7 pages. 6 figures. For source code, refer https://github.com/madhawav/MML
HIGHLIGHT:      We propose "Multi-faceted VideoMoment Localizer" (MML), an extension of MAC model by the introduction of visual object evidence via object segmentation masks and video understanding features via video captioning.

22, TITLE:      Multi-branch Attentive Transformer
http://arxiv.org/abs/2006.10270
AUTHORS:        Yang Fan ; Shufang Xie ; Yingce Xia ; Lijun Wu ; Tao Qin ; Xiang-Yang Li ; Tie-Yan Liu
COMMENTS:       17 pages
HIGHLIGHT:      In this work, we propose a simple yet effective variant of Transformer called multi-branch attentive Transformer (briefly, MAT), where the attention layer is the average of multiple branches and each branch is an independent multi-head attention layer.

23, TITLE:      Compositional theories for embedded languages
http://arxiv.org/abs/2006.10604
AUTHORS:        Davide Trotta ; Margherita Zorzi
COMMENTS:       20 pages
HIGHLIGHT:      The aim of this paper is to present a flexible fragment of such a type theory, together with its categorical semantics in terms of enriched categories, following previous investigations.

24, TITLE:      The Clever Hans Effect in Anomaly Detection
http://arxiv.org/abs/2006.10609
AUTHORS:        Jacob Kauffmann ; Lukas Ruff ; Grégoire Montavon ; Klaus-Robert Müller
COMMENTS:       17 pages, preprint
HIGHLIGHT:      Therefore, this paper will contribute an explainable AI (XAI) procedure that can highlight the relevant features used by popular anomaly detection models of different type.

25, TITLE:      SatImNet: Structured and Harmonised Training Data for Enhanced Satellite Imagery Classification
http://arxiv.org/abs/2006.10623
AUTHORS:        Vasileios Syrris ; Ondrej Pesek ; Pierre Soille
HIGHLIGHT:      To alleviate these problems, we propose a methodology for structuring and harmonising open training datasets on the basis of a series of fundamental attributes we put forward for any such dataset.

26, TITLE:      On the Predictability of Pruning Across Scales
http://arxiv.org/abs/2006.10621
AUTHORS:        Jonathan S. Rosenfeld ; Jonathan Frankle ; Michael Carbin ; Nir Shavit
HIGHLIGHT:      We show that the error of magnitude-pruned networks follows a scaling law, and that this law is of a fundamentally different nature than that of unpruned networks.

27, TITLE:      Use of in-the-wild images for anomaly detection in face anti-spoofing
http://arxiv.org/abs/2006.10626
AUTHORS:        Latifah Abduh ; Ioannis Ivrissimtzis
COMMENTS:       6 pages
HIGHLIGHT:      Here, we explore the use of in-the-wild images, and images from non-specialized face databases, to train one-class classifiers for face anti-spoofing.

28, TITLE:      Compositional Generalization by Learning Analytical Expressions
http://arxiv.org/abs/2006.10627
AUTHORS:        Qian Liu ; Shengnan An ; Jian-Guang Lou ; Bei Chen ; Zeqi Lin ; Yan Gao ; Bin Zhou ; Nanning Zheng ; Dongmei Zhang
COMMENTS:       14 pages, 6 figures
HIGHLIGHT:      Inspired by work in cognition which argues compositionality can be captured by variable slots with symbolic functions, we present a refreshing view that connects a memory-augmented neural model with analytical expressions, to achieve compositional generalization.

29, TITLE:      Explainable and Discourse Topic-aware Neural Language Understanding
http://arxiv.org/abs/2006.10632
AUTHORS:        Yatin Chaudhary ; Hinrich Schütze ; Pankaj Gupta
COMMENTS:       Accepted at ICML2020 (13 pages, 2 figures)
HIGHLIGHT:      We present a novel neural composite language model that exploits both the latent and explainable topics along with topical discourse at sentence-level in a joint learning framework of topic and language models.

30, TITLE:      Online Deep Clustering for Unsupervised Representation Learning
http://arxiv.org/abs/2006.10645
AUTHORS:        Xiaohang Zhan ; Jiahao Xie ; Ziwei Liu ; Yew Soon Ong ; Chen Change Loy
COMMENTS:       Accepted by CVPR 2020. Code: https://github.com/open-mmlab/OpenSelfSup
HIGHLIGHT:      To overcome this challenge, we propose Online Deep Clustering (ODC) that performs clustering and network update simultaneously rather than alternatingly.

31, TITLE:      Fourth-Order Anisotropic Diffusion for Inpainting and Image Compression
http://arxiv.org/abs/2006.10406
AUTHORS:        Ikram Jumakulyyev ; Thomas Schultz
COMMENTS:       Accepted for publication in Springer book "Anisotropy Across Fields and Scales"
HIGHLIGHT:      In this work, we generalize second-order EED to a fourth-order counterpart.

32, TITLE:      Multi-Density Sketch-to-Image Translation Network
http://arxiv.org/abs/2006.10649
AUTHORS:        Jialu Huang ; Jing Liao ; Zhifeng Tan ; Sam Kwong
COMMENTS:       2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
HIGHLIGHT:      In this work, we propose the first multi-level density sketch-to-image translation framework, which allows the input sketch to cover a wide range from rough object outlines to micro structures.

33, TITLE:      Overcoming Classifier Imbalance for Long-tail Object Detection with Balanced Group Softmax
http://arxiv.org/abs/2006.10408
AUTHORS:        Yu Li ; Tao Wang ; Bingyi Kang ; Sheng Tang ; Chunfeng Wang ; Jintao Li ; Jiashi Feng
COMMENTS:       CVPR 2020 (Oral). Code is available at https://github.com/FishYuLi/BalancedGroupSoftmax
HIGHLIGHT:      In this work, we provide the first systematic analysis on the underperformance of state-of-the-art models in front of long-tail distribution.

34, TITLE:      Pre-trained Language Models as Symbolic Reasoners over Knowledge?
http://arxiv.org/abs/2006.10413
AUTHORS:        Nora Kassner ; Benno Kroje ; Hinrich Schütze
COMMENTS:       work in progress
HIGHLIGHT:      Prior work has attempted to quantify the number of facts PLMs learn, but we present, using synthetic data, the first study that establishes a causal relation between facts present in training and facts learned by the PLM.

35, TITLE:      Toric Eigenvalue Methods for Solving Sparse Polynomial Systems
http://arxiv.org/abs/2006.10654
AUTHORS:        Matías R. Bender ; Simon Telen
COMMENTS:       41 pages, 7 figures
HIGHLIGHT:      In this work, we give a first description of this strategy for non-reduced, zero-dimensional subschemes of $X$.

36, TITLE:      Is this Dialogue Coherent? Learning from Dialogue Acts and Entities
http://arxiv.org/abs/2006.10157
AUTHORS:        Alessandra Cervone ; Giuseppe Riccardi
COMMENTS:       Accepted at SIGDIAL 2020
HIGHLIGHT:      In this work, we investigate the human perception of coherence in open-domain dialogues. First, we create the Switchboard Coherence (SWBD-Coh) corpus, a dataset of human-human spoken dialogues annotated with turn coherence ratings, where next-turn candidate utterances ratings are provided considering the full dialogue context.

37, TITLE:      Deep Network for Scatterer Distribution Estimation for Ultrasound Image Simulation
http://arxiv.org/abs/2006.10166
AUTHORS:        Lin Zhang ; Valery Vishnevskiy ; Orcun Goksel
HIGHLIGHT:      In this paper, we demonstrate a convolutional neural network approach for probabilistic scatterer estimation from observed ultrasound data.

38, TITLE:      MIMICS: A Large-Scale Data Collection for Search Clarification
http://arxiv.org/abs/2006.10174
AUTHORS:        Hamed Zamani ; Gord Lueck ; Everest Chen ; Rodolfo Quispe ; Flint Luu ; Nick Craswell
HIGHLIGHT:      In this paper, we introduce MIMICS, a collection of search clarification datasets for real web search queries sampled from the Bing query logs.

39, TITLE:      Variational State-Space Models for Localisation and Dense 3D Mapping in 6 DoF
http://arxiv.org/abs/2006.10178
AUTHORS:        Atanas Mirchev ; Baris Kayalibay ; Patrick van der Smagt ; Justin Bayer
HIGHLIGHT:      We solve the problem of 6-DoF localisation and 3D dense reconstruction in spatial environments as approximate Bayesian inference in a deep generative approach which combines learned with engineered models.

40, TITLE:      Stochastic Bandits with Linear Constraints
http://arxiv.org/abs/2006.10185
AUTHORS:        Aldo Pacchiano ; Mohammad Ghavamzadeh ; Peter Bartlett ; Heinrich Jiang
COMMENTS:       9 pages
HIGHLIGHT:      We further specialize our results to multi-armed bandits and propose a computationally efficient algorithm for this setting.

41, TITLE:      TearingNet: Point Cloud Autoencoder to Learn Topology-Friendly Representations
http://arxiv.org/abs/2006.10187
AUTHORS:        Jiahao Pang ; Duanshun Li ; Dong Tian
COMMENTS:       Submitted to NeurIPS 2020
HIGHLIGHT:      Given a point cloud dataset containing objects with various genera or scenes with multiple objects, we propose an autoencoder, TearingNet, which tackles the challenging task of representing the point clouds using a fixed-length descriptor.

42, TITLE:      Head2Head++: Deep Facial Attributes Re-Targeting
http://arxiv.org/abs/2006.10199
AUTHORS:        Michail Christos Doukas ; Mohammad Rami Koujan ; Viktoriia Sharmanska ; Anastasios Roussos
COMMENTS:       Submitted to the IEEE Transactions on Biometrics, Behavior, and Identity Science (TBIOM) journal
HIGHLIGHT:      We leverage the 3D geometry of faces and Generative Adversarial Networks (GANs) to design a novel deep learning architecture for the task of facial and head reenactment.

43, TITLE:      Neural Graphics Pipeline for Controllable Image Generation
http://arxiv.org/abs/2006.10569
AUTHORS:        Xuelin Chen ; Daniel Cohen-Or ; Baoquan Chen ; Niloy J. Mitra
COMMENTS:       15 pages, 10 figures
HIGHLIGHT:      We present Neural Graphics Pipeline (NGP), a hybrid generative model that brings together neural and traditional image formation models.

44, TITLE:      A Review of 1D Convolutional Neural Networks toward Unknown Substance Identification in Portable Raman Spectrometer
http://arxiv.org/abs/2006.10575
AUTHORS:        M. Hamed Mozaffari ; Li-Lin Tay
COMMENTS:       19 pages, 1 figure, 5 tables
HIGHLIGHT:      In this study, we present a comprehensive survey in the use of one-dimensional CNNs for Raman spectrum identification.

45, TITLE:      A Shooting Formulation of Deep Learning
http://arxiv.org/abs/2006.10330
AUTHORS:        François-Xavier Vialard ; Roland Kwitt ; Susan Wei ; Marc Niethammer
HIGHLIGHT:      To this end, we introduce a shooting formulation which shifts the perspective from parameterizing a network layer-by-layer to parameterizing over optimal networks described only by a set of initial conditions.

46, TITLE:      Cascaded Regression Tracking: Towards Online Hard Distractor Discrimination
http://arxiv.org/abs/2006.10336
AUTHORS:        Ning Wang ; Wengang Zhou ; Qi Tian ; Houqiang Li
COMMENTS:       Accepted by IEEE TCSVT
HIGHLIGHT:      To enhance the tracking robustness, in this paper, we propose a cascaded regression tracker with two sequential stages.

47, TITLE:      Extraction and Evaluation of Formulaic Expressions Used in Scholarly Papers
http://arxiv.org/abs/2006.10334
AUTHORS:        Kenichi Iwatsuki ; Florian Boudin ; Akiko Aizawa
COMMENTS:       21 pages, 11 figures
HIGHLIGHT:      In this paper, we propose a new approach that is robust to variation of spans and forms of formulaic expressions.

48, TITLE:      Rethinking Semi-Supervised Learning in VAEs
http://arxiv.org/abs/2006.10102
AUTHORS:        Tom Joy ; Sebastian M. Schmon ; Philip H. S. Torr ; N. Siddharth ; Tom Rainforth
HIGHLIGHT:      We present an alternative approach to semi-supervision in variational autoencoders(VAEs) that incorporates labels through auxiliary variables rather than directly through the latent variables.

49, TITLE:      Automated Radiological Report Generation For Chest X-Rays With Weakly-Supervised End-to-End Deep Learning
http://arxiv.org/abs/2006.10347
AUTHORS:        Shuai Zhang ; Xiaoyan Xin ; Yang Wang ; Yachong Guo ; Qiuqiao Hao ; Xianfeng Yang ; Jun Wang ; Jian Zhang ; Bing Zhang ; Wei Wang
HIGHLIGHT:      In this work, we built a database containing more than 12,000 CXR scans and radiological reports, and developed a model based on deep convolutional neural network and recurrent network with attention mechanism.

50, TITLE:      Is Network the Bottleneck of Distributed Training?
http://arxiv.org/abs/2006.10103
AUTHORS:        Zhen Zhang ; Chaokun Chang ; Haibin Lin ; Yida Wang ; Raman Arora ; Xin Jin
HIGHLIGHT:      In this paper, we take a first-principles approach to measure and analyze the network performance of distributed training.

51, TITLE:      On the complexity of detecting hazards
http://arxiv.org/abs/2006.10592
AUTHORS:        Balagopal Komarath ; Nitin Saurabh
COMMENTS:       To appear in Information Processing Letters
HIGHLIGHT:      We show that there is no $O(3^{(1-\epsilon)n} \text{poly}(s))$ time algorithm, for any $\epsilon > 0$, that detects logic hazards in Boolean circuits of size $s$ on $n$ variables under the assumption that the strong exponential time hypothesis is true.

52, TITLE:      Shapeshifter Networks: Cross-layer Parameter Sharing for Scalable and Effective Deep Learning
http://arxiv.org/abs/2006.10598
AUTHORS:        Bryan A. Plummer ; Nikoli Dryden ; Julius Frost ; Torsten Hoefler ; Kate Saenko
HIGHLIGHT:      We present Shapeshifter Networks (SSNs), a flexible neural network framework that improves performance and reduces memory requirements on a diverse set of scenarios over standard neural networks.

53, TITLE:      Learning by Repetition: Stochastic Multi-armed Bandits under Priming Effect
http://arxiv.org/abs/2006.10356
AUTHORS:        Priyank Agrawal ; Theja Tulabandhula
COMMENTS:       Appears in the 36th Conference on Uncertainty in Artificial Intelligence (UAI 2020)
HIGHLIGHT:      We provide novel algorithms that achieves sublinear regret in time and the relevant wear-in/wear-out parameters.

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

55, TITLE:      On the Robustness of Active Learning
http://arxiv.org/abs/2006.10370
AUTHORS:        Lukas Hahn ; Lutz Roese-Koerner ; Peet Cremer ; Urs Zimmermann ; Ori Maoz ; Anton Kummert
COMMENTS:       11 pages, 6 figures, 1 table; as published in the proceedings of the 5th Global Conference on Artificial Intelligence (GCAI), EPiC Series in Computing, Volume 65, pages 152-162, https://doi.org/10.29007/thws, 2019
HIGHLIGHT:      In this work we analyse the robustness of different Active Learning methods with respect to classifier capacity, exchangeability and type, as well as hyperparameters and falsely labelled data.

56, TITLE:      A Practical Online Method for Distributionally Deep Robust Optimization
http://arxiv.org/abs/2006.10138
AUTHORS:        Qi Qi ; Zhishuai Guo ; Yi Xu ; Rong Jin ; Tianbao Yang
COMMENTS:       20 pages, 8 figures
HIGHLIGHT:      In this paper, we propose a practical online method for solving a distributionally robust optimization (DRO) for deep learning, which has important applications in machine learning for improving the robustness of neural networks.

57, TITLE:      Video Semantic Segmentation with Distortion-Aware Feature Correction
http://arxiv.org/abs/2006.10380
AUTHORS:        Jiafan Zhuang ; Zilei Wang ; Bingke Wang
HIGHLIGHT:      In this paper, we propose distortion-aware feature correction to alleviate the issue, which improves video segmentation performance by correcting distorted propagated features.

58, TITLE:      3D Pipe Network Reconstruction Based on Structure from Motion with Incremental Conic Shape Detection and Cylindrical Constraint
http://arxiv.org/abs/2006.10383
AUTHORS:        Sho kagami ; Hajime Taira ; Naoyuki Miyashita ; Akihiko Torii ; Masatoshi Okutomi
HIGHLIGHT:      In this paper, we propose a 3D pipe reconstruction system using sequential images captured by a monocular endoscopic camera.

59, TITLE:      Using Sentiment Information for Preemptive Detection of Toxic Comments in Online Conversations
http://arxiv.org/abs/2006.10145
AUTHORS:        Éloi Brassard-Gourdeau ; Richard Khoury
HIGHLIGHT:      In this paper, we combine that approach with previous work on toxicity detection using sentiment information, and show how the sentiments expressed in the first messages of a conversation can help predict upcoming toxicity.

60, TITLE:      SceneAdapt: Scene-based domain adaptation for semantic segmentation using adversarial learning
http://arxiv.org/abs/2006.10386
AUTHORS:        Daniele Di Mauro ; Antonino Furnari ; Giuseppe Patanè ; Sebastiano Battiato ; Giovanni Maria Farinella
HIGHLIGHT:      As a first approach to address this challenging task, we propose SceneAdapt, a method for scene adaptation of semantic segmentation algorithms based on adversarial learning. We formalize this problem as a domain adaptation task and introduce a novel dataset of urban scenes with the related semantic labels.

61, TITLE:      Are you wearing a mask? Improving mask detection from speech using augmentation by cycle-consistent GANs
http://arxiv.org/abs/2006.10147
AUTHORS:        Nicolae-Cătălin Ristea ; Radu Tudor Ionescu
HIGHLIGHT:      In this paper, we propose a novel data augmentation approach for mask detection from speech.

62, TITLE:      Unsupervised out-of-distribution detection using kernel density estimation
http://arxiv.org/abs/2006.10712
AUTHORS:        Ertunc Erdil ; Krishna Chaitanya ; Ender Konukoglu
HIGHLIGHT:      In this paper, we propose an unsupervised OOD detection method that can work with both classification and non-classification networks by using kernel density estimation (KDE).

63, TITLE:      Zero-Shot Learning with Common Sense Knowledge Graphs
http://arxiv.org/abs/2006.10713
AUTHORS:        Nihal V. Nayak ; Stephen H. Bach
HIGHLIGHT:      To capture the knowledge in the graph, we introduce ZSL-KG, a framework based on graph neural networks with non-linear aggregators to generate class representations.

64, TITLE:      Ocean: Object-aware Anchor-free Tracking
http://arxiv.org/abs/2006.10721
AUTHORS:        Zhipeng Zhang ; Houwen Peng
HIGHLIGHT:      In this paper, we propose a novel object-aware anchor-free network to address this issue.

65, TITLE:      IReEn: Iterative Reverse-Engineering of Black-Box Functions via Neural Program Synthesis
http://arxiv.org/abs/2006.10720
AUTHORS:        Hossein Hajipour ; Mateusz Malinowski ; Mario Fritz
COMMENTS:       14 pages, 10 figures
HIGHLIGHT:      In this work, we investigate the problem of revealing the functionality of a black-box agent.

66, TITLE:      Fully Test-time Adaptation by Entropy Minimization
http://arxiv.org/abs/2006.10726
AUTHORS:        Dequan Wang ; Evan Shelhamer ; Shaoteng Liu ; Bruno Olshausen ; Trevor Darrell
HIGHLIGHT:      We propose an entropy minimization approach for adaptation: we take the model's confidence as our objective as measured by the entropy of its predictions.

67, TITLE:      Cyclic Differentiable Architecture Search
http://arxiv.org/abs/2006.10724
AUTHORS:        Hongyuan Yu ; Houwen Peng
HIGHLIGHT:      To address this issue, we propose a novel cyclic differentiable architecture search framework (CDARTS).

68, TITLE:      Diverse Image Generation via Self-Conditioned GANs
http://arxiv.org/abs/2006.10728
AUTHORS:        Steven Liu ; Tongzhou Wang ; David Bau ; Jun-Yan Zhu ; Antonio Torralba
COMMENTS:       CVPR 2020. Code: https://github.com/stevliu/self-conditioned-gan. Webpage: http://selfcondgan.csail.mit.edu/
HIGHLIGHT:      We introduce a simple but effective unsupervised method for generating realistic and diverse images.

69, TITLE:      A unified framework for equivalences in social networks
http://arxiv.org/abs/2006.10733
AUTHORS:        Nina Otter ; Mason A. Porter
COMMENTS:       working paper
HIGHLIGHT:      Motivated by the principle of functoriality in category theory we propose a new method that allows to tie role and positional analysis together.

70, TITLE:      Spin-Weighted Spherical CNNs
http://arxiv.org/abs/2006.10731
AUTHORS:        Carlos Esteves ; Ameesh Makadia ; Kostas Daniilidis
HIGHLIGHT:      In this paper, we present a new type of spherical CNN that allows anisotropic filters in an efficient way, without ever leaving the spherical domain.

71, TITLE:      Forward Prediction for Physical Reasoning
http://arxiv.org/abs/2006.10734
AUTHORS:        Rohit Girdhar ; Laura Gustafson ; Aaron Adcock ; Laurens van der Maaten
HIGHLIGHT:      We do so by incorporating models that operate on object or pixel-based representations of the world, into simple physical-reasoning agents.

72, TITLE:      Differentiable Augmentation for Data-Efficient GAN Training
http://arxiv.org/abs/2006.10738
AUTHORS:        Shengyu Zhao ; Zhijian Liu ; Ji Lin ; Jun-Yan Zhu ; Song Han
HIGHLIGHT:      To combat it, we propose Differentiable Augmentation (DiffAugment), a simple method that improves the data efficiency of GANs by imposing various types of differentiable augmentations on both real and fake samples.

73, TITLE:      Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
http://arxiv.org/abs/2006.10739
AUTHORS:        Matthew Tancik ; Pratul P. Srinivasan ; Ben Mildenhall ; Sara Fridovich-Keil ; Nithin Raghavan ; Utkarsh Singhal ; Ravi Ramamoorthi ; Jonathan T. Barron ; Ren Ng
COMMENTS:       Project page: https://people.eecs.berkeley.edu/~bmild/fourfeat/
HIGHLIGHT:      We suggest an approach for selecting problem-specific Fourier features that greatly improves the performance of MLPs for low-dimensional regression tasks relevant to the computer vision and graphics communities.

74, TITLE:      Learning Invariant Representations for Reinforcement Learning without Reconstruction
http://arxiv.org/abs/2006.10742
AUTHORS:        Amy Zhang ; Rowan McAllister ; Roberto Calandra ; Yarin Gal ; Sergey Levine
HIGHLIGHT:      Our goal is to learn representations that both provide for effective downstream control and invariance to task-irrelevant details.

75, TITLE:      Practical Large-Scale Distributed Parallel Monte-Carlo Tree Search Applied to Molecular Design
http://arxiv.org/abs/2006.10504
AUTHORS:        Xiufeng Yang ; Tanuj Kr Aasawat ; Kazuki Yoshizoe
HIGHLIGHT:      In this paper, we propose to apply a hash function based distributed parallel Monte-Carlo Tree Search (MCTS) to a real-world problem of molecular design.

76, TITLE:      Structure and Design of HoloGen
http://arxiv.org/abs/2006.10509
AUTHORS:        Peter J. Christopher ; Timothy D. Wilkinson
HIGHLIGHT:      This article discusses the structure and design of HoloGen.

77, TITLE:      Contrastive learning of global and local features for medical image segmentation with limited annotations
http://arxiv.org/abs/2006.10511
AUTHORS:        Krishna Chaitanya ; Ertunc Erdil ; Neerav Karani ; Ender Konukoglu
COMMENTS:       16 pages, 2 figures, 7 tables. This article is a pre-print and is currently under review at a conference
HIGHLIGHT:      In this work, we propose strategies for extending the contrastive learning framework for segmentation of volumetric medical images in the semi-supervised setting with limited annotations, by leveraging domain-specific and problem-specific cues.

78, TITLE:      Automatic Speech Recognition Benchmark for Air-Traffic Communications
http://arxiv.org/abs/2006.10304
AUTHORS:        Juan Zuluaga-Gomez ; Petr Motlicek ; Qingran Zhan ; Karel Vesely ; Rudolf Braun
COMMENTS:       Submitted to the 21st INTERSPEECH conference (Shanghai, October 25-29)
HIGHLIGHT:      Hereby, we introduce CleanSky EC-H2020 ATCO2, a project that aims to develop an ASR-based platform to collect, organize and automatically pre-process ATCo speech-data from air space.

79, TITLE:      Latent Video Transformer
http://arxiv.org/abs/2006.10704
AUTHORS:        Ruslan Rakhimov ; Denis Volkhonskiy ; Alexey Artemov ; Denis Zorin ; Evgeny Burnaev
HIGHLIGHT:      In this work, we address this problem via modeling the dynamics in a latent space.

80, TITLE:      Deep Reinforcement Learning amidst Lifelong Non-Stationarity
http://arxiv.org/abs/2006.10701
AUTHORS:        Annie Xie ; James Harrison ; Chelsea Finn
COMMENTS:       supplementary website at https://sites.google.com/stanford.edu/lilac/
HIGHLIGHT:      As humans, our goals and our environment are persistently changing throughout our lifetime based on our experiences, actions, and internal and external drives.

81, TITLE:      Semi-Supervised Recognition under a Noisy and Fine-grained Dataset
http://arxiv.org/abs/2006.10702
AUTHORS:        Cheng Cui ; Zhi Ye ; Yangxi Li ; Xinjian Li ; Min Yang ; Kai Wei ; Bing Dai ; Yanmei Zhao ; Zhongji Liu ; Rong Pang
COMMENTS:       5 pages, 3 figures, 3 tables
HIGHLIGHT:      We combined generic image recognition and fine-grained image recognition method to solve the problem.

82, TITLE:      Set Distribution Networks: a Generative Model for Sets of Images
http://arxiv.org/abs/2006.10705
AUTHORS:        Shuangfei Zhai ; Walter Talbott ; Miguel Angel Bautista ; Carlos Guestrin ; Josh M. Susskind
HIGHLIGHT:      We introduce Set Distribution Networks (SDNs), a novel framework that learns to autoencode and freely generate sets.

83, TITLE:      Octet: Online Catalog Taxonomy Enrichment with Self-Supervision
http://arxiv.org/abs/2006.10276
AUTHORS:        Yuning Mao ; Tong Zhao ; Andrey Kan ; Chenwei Zhang ; Xin Luna Dong ; Christos Faloutsos ; Jiawei Han
COMMENTS:       KDD 2020
HIGHLIGHT:      In this paper, we present a self-supervised end-to-end framework, Octet, for Online Catalog Taxonomy EnrichmenT.

84, TITLE:      Joint Contrastive Learning for Unsupervised Domain Adaptation
http://arxiv.org/abs/2006.10297
AUTHORS:        Changhwa Park ; Jonghyun Lee ; Jaeyoon Yoo ; Minhoe Hur ; Sungroh Yoon
COMMENTS:       16 pages, 1 figure, 4 tables
HIGHLIGHT:      In this paper, we propose an alternative upper bound on the target error that explicitly considers the joint error to render it more manageable.

85, TITLE:      Overcoming Statistical Shortcuts for Open-ended Visual Counting
http://arxiv.org/abs/2006.10079
AUTHORS:        Corentin Dancette ; Remi Cadene ; Xinlei Chen ; Matthieu Cord
COMMENTS:       17 pages, 8 figures
HIGHLIGHT:      We aim to develop models that learn a proper mechanism of counting regardless of the output label.

86, TITLE:      Fair k-Means Clustering
http://arxiv.org/abs/2006.10085
AUTHORS:        Mehrdad Ghadiri ; Samira Samadi ; Santosh Vempala
COMMENTS:       17 pages, 9 figures
HIGHLIGHT:      We present a fair $k$-means objective and algorithm to choose cluster centers that provide equitable costs for different groups.

87, TITLE:      Extensively Matching for Few-shot Learning Event Detection
http://arxiv.org/abs/2006.10093
AUTHORS:        Viet Dac Lai ; Franck Dernoncourt ; Thien Huu Nguyen
COMMENTS:       1st Joint Workshop on Narrative Understanding, Storylines, and Events (NUSE) @ ACL 2020
HIGHLIGHT:      In this work, weformulate event detection as a few-shot learn-ing problem to enable to extend event detec-tion to new event types.

88, TITLE:      Faster Secure Data Mining via Distributed Homomorphic Encryption
http://arxiv.org/abs/2006.10091
AUTHORS:        Junyi Li ; Heng Huang
HIGHLIGHT:      In this paper, we propose a novel general distributed HE-based data mining framework towards one step of solving the scaling problem.

89, TITLE:      Towards Recurrent Autoregressive Flow Models
http://arxiv.org/abs/2006.10096
AUTHORS:        John Mern ; Peter Morales ; Mykel J. Kochenderfer
HIGHLIGHT:      In this work, we present an initial design for a recurrent flow cell and a method to train the model to match observed empirical distributions.



==========Updates to Previous Papers==========
1, TITLE:       An initial attempt of combining visual selective attention with deep reinforcement learning
http://arxiv.org/abs/1811.04407
AUTHORS:        Liu Yuezhang ; Ruohan Zhang ; Dana H. Ballard
COMMENTS:       7 pages, 8 figures
HIGHLIGHT:      We visualize and analyze the feature maps of DQN on a toy problem Catch, and propose an approach to combine visual selective attention with deep reinforcement learning.

2, TITLE:       On the Generation of Medical Dialogues for COVID-19
http://arxiv.org/abs/2005.05442
AUTHORS:        Wenmian Yang ; Guangtao Zeng ; Bowen Tan ; Zeqian Ju ; Subrato Chakravorty ; Xuehai He ; Shu Chen ; Xingyi Yang ; Qingyang Wu ; Zhou Yu ; Eric Xing ; Pengtao Xie
HIGHLIGHT:      To address this problem, we aim to develop a medical dialogue system that can provide COVID19-related consultations. We collected two dialogue datasets -- CovidDialog -- (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19.

3, TITLE:       Deep Energy-Based Modeling of Discrete-Time Physics
http://arxiv.org/abs/1905.08604
AUTHORS:        Takashi Matsubara ; Ai Ishikawa ; Takaharu Yaguchi
COMMENTS:       This is a thorough revision that focuses on applications to data-driven modeling
HIGHLIGHT:      In this study, we propose a deep energy-based physical model that admits a specific differential geometric structure.

4, 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.

5, 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.

6, TITLE:       An Iterative Approach for Identifying Complaint Based Tweets in Social Media Platforms
http://arxiv.org/abs/2001.09215
AUTHORS:        Gyanesh Anand ; Akash Gautam ; Puneet Mathur ; Debanjan Mahata ; Rajiv Ratn Shah ; Ramit Sawhney
COMMENTS:       Preprint of paper accepted at AAAI, student abstract 2020
HIGHLIGHT:      In this paper, we propose an iterative methodology which aims to identify complaint based posts pertaining to the transport domain.

7, TITLE:       Automatic Validation of Textual Attribute Values in E-commerce Catalog by Learning with Limited Labeled Data
http://arxiv.org/abs/2006.08779
AUTHORS:        Yaqing Wang ; Yifan Ethan Xu ; Xian Li ; Xin Luna Dong ; Jing Gao
COMMENTS:       Proceedings of the 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, August 23--27, 2020, Virtual Event, CA, USA
HIGHLIGHT:      To address the aforementioned challenges, we propose a novel meta-learning latent variable approach, called MetaBridge, which can learn transferable knowledge from a subset of categories with limited labeled data and capture the uncertainty of never-seen categories with unlabeled data.

8, 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.

9, TITLE:       No-regret Exploration in Contextual Reinforcement Learning
http://arxiv.org/abs/1903.06187
AUTHORS:        Aditya Modi ; Ambuj Tewari
COMMENTS:       Accepted to UAI 2020. PMLR proceedings, volume 124
HIGHLIGHT:      In this paper, we propose a no-regret online RL algorithm in the setting where the MDP parameters are obtained from the context using generalized linear mappings (GLMs).

10, TITLE:      ActiveMoCap: Optimized Viewpoint Selection for Active Human Motion Capture
http://arxiv.org/abs/1912.08568
AUTHORS:        Sena Kiciroglu ; Helge Rhodin ; Sudipta N. Sinha ; Mathieu Salzmann ; Pascal Fua
COMMENTS:       For associated video, see https://youtu.be/i58Bu-hbZHs Published in CVPR 2020
HIGHLIGHT:      Specifically, given a short video sequence, we introduce an algorithm that predicts which viewpoints should be chosen to capture future frames so as to maximize 3D human pose estimation accuracy.

11, TITLE:      Batch Normalization Biases Residual Blocks Towards the Identity Function in Deep Networks
http://arxiv.org/abs/2002.10444
AUTHORS:        Soham De ; Samuel L. Smith
HIGHLIGHT:      We show that this key benefit arises because, at initialization, batch normalization downscales the residual branch relative to the skip connection, by a normalizing factor on the order of the square root of the network depth.

12, TITLE:      Automated Thalamic Nuclei Segmentation Using Multi-Planar Cascaded Convolutional Neural Networks
http://arxiv.org/abs/1912.07209
AUTHORS:        Mohammad S Majdi ; Mahesh B Keerthivasan ; Brian K Rutt ; Natalie M Zahr ; Jeffrey J Rodriguez ; Manojkumar Saranathan
COMMENTS:       Submitted to Magnetic Resonance Imaging. 34 pages, 6 figures , 2 tables, 1 supporting figures, 2 supporting tables
HIGHLIGHT:      Clinical utility was demonstrated by applying this method to study the effect of MS on thalamic nuclei atrophy.

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:      COVID-CT-Dataset: A CT Scan Dataset about COVID-19
http://arxiv.org/abs/2003.13865
AUTHORS:        Xingyi Yang ; Xuehai He ; Jinyu Zhao ; Yichen Zhang ; Shanghang Zhang ; Pengtao Xie
HIGHLIGHT:      Using this dataset, we develop diagnosis methods based on multi-task learning and self-supervised learning, that achieve an F1 of 0.90, an AUC of 0.98, and an accuracy of 0.89. To address this issue, we build an open-sourced dataset -- COVID-CT, which contains 349 COVID-19 CT images from 216 patients and 463 non-COVID-19 CTs.

15, TITLE:      Probabilistic Reasoning across the Causal Hierarchy
http://arxiv.org/abs/2001.02889
AUTHORS:        Duligur Ibeling ; Thomas Icard
COMMENTS:       AAAI-20
HIGHLIGHT:      We propose a formalization of the three-tier causal hierarchy of association, intervention, and counterfactuals as a series of probabilistic logical languages.

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:      CERT: Contrastive Self-supervised Learning for Language Understanding
http://arxiv.org/abs/2005.12766
AUTHORS:        Hongchao Fang ; Sicheng Wang ; Meng Zhou ; Jiayuan Ding ; Pengtao Xie
HIGHLIGHT:      To address this issue, we propose CERT: Contrastive self-supervised Encoder Representations from Transformers, which pretrains language representation models using contrastive self-supervised learning at the sentence level.

18, TITLE:      Action Anticipation for Collaborative Environments: The Impact of Contextual Information and Uncertainty-Based Prediction
http://arxiv.org/abs/1910.00714
AUTHORS:        Clebeson Canuto ; Plinio Moreno ; Jorge Samatelo ; Raquel Vassallo ; José Santos-Victor
COMMENTS:       27 pages, 16 figures, Neurocomputing
HIGHLIGHT:      In this work, we consider two additional sources of information (i.e., context) over time, gaze, movement and object information, and study how these additional contextual cues improve the action anticipation performance.

19, TITLE:      Constructing Parsimonious Analytic Models for Dynamic Systems via Symbolic Regression
http://arxiv.org/abs/1903.11483
AUTHORS:        Erik Derner ; Jiří Kubalík ; Nicola Ancona ; Robert Babuška
HIGHLIGHT:      In this paper, we propose to employ symbolic regression (SR) to construct parsimonious process models described by analytic equations.

20, TITLE:      Distinguishing noisy boson sampling from classical simulations
http://arxiv.org/abs/1905.11458
AUTHORS:        Valery Shchesnovich
COMMENTS:       18 pages (6 pages of the main text), two figures (both in colour)
HIGHLIGHT:      In this work it is shown that one can efficiently distinguish the output distribution of such a noisy boson sampling from any approximation accounting for the low-order quantum multiboson interferences, which includes the mentioned classical algorithms.

21, TITLE:      The efficiency of deep learning algorithms for detecting anatomical reference points on radiological images of the head profile
http://arxiv.org/abs/2005.12110
AUTHORS:        Konstantin Dobratulin ; Andrey Gaidel ; Irina Aupova ; Anna Ivleva ; Aleksandr Kapishnikov ; Pavel Zelter
HIGHLIGHT:      In this article we investigate the efficiency of deep learning algorithms in solving the task of detecting anatomical reference points on radiological images of the head in lateral projection using a fully convolutional neural network and a fully convolutional neural network with an extended architecture for biomedical image segmentation - U-Net.

22, TITLE:      Biometric Quality: Review and Application to Face Recognition with FaceQnet
http://arxiv.org/abs/2006.03298
AUTHORS:        Javier Hernandez-Ortega ; Javier Galbally ; Julian Fierrez ; Laurent Beslay
HIGHLIGHT:      After a gentle introduction to the general topic of biometric quality and a review of past efforts in face quality metrics, in the present work, we address the need for better face quality metrics by developing FaceQnet.

23, TITLE:      PoWER-BERT: Accelerating BERT Inference via Progressive Word-vector Elimination
http://arxiv.org/abs/2001.08950
AUTHORS:        Saurabh Goyal ; Anamitra R. Choudhary ; Saurabh M. Raje ; Venkatesan Chakaravarthy ; Yogish Sabharwal ; Ashish Verma
COMMENTS:       11 pages, 8 figures, 4 tables
HIGHLIGHT:      We develop a novel method, called PoWER-BERT, for improving the inference time of the popular BERT model, while maintaining the accuracy.

24, TITLE:      Burst Denoising of Dark Images
http://arxiv.org/abs/2003.07823
AUTHORS:        Ahmet Serdar Karadeniz ; Erkut Erdem ; Aykut Erdem
COMMENTS:       This paper has been withdrawn by the authors to be replaced by a new version available at arXiv:2006.09845
HIGHLIGHT:      Motivated by these ideas, in this paper, we propose a deep learning framework for obtaining clean and colorful RGB images from extremely dark raw images.

25, TITLE:      Uncertainty in Structured Prediction
http://arxiv.org/abs/2002.07650
AUTHORS:        Andrey Malinin ; Mark Gales
HIGHLIGHT:      Thus, this work aims to investigate uncertainty estimation for structured prediction tasks within a single unified and interpretable probabilistic ensemble-based framework.

26, TITLE:      Dispersed EM-VAEs for Interpretable Text Generation
http://arxiv.org/abs/1906.06719
AUTHORS:        Wenxian Shi ; Hao Zhou ; Ning Miao ; Lei Li
COMMENTS:       Accepted by ICML 2020
HIGHLIGHT:      In this paper, we find that mode-collapse is a general problem for VAEs with exponential family mixture priors.

27, TITLE:      A Simple and Effective Framework for Pairwise Deep Metric Learning
http://arxiv.org/abs/1912.11194
AUTHORS:        Qi Qi ; Yan Yan ; Xiaoyu Wang ; Tianbao Yang
COMMENTS:       16 pages, 5 figures
HIGHLIGHT:      In this paper, we cast DML as a simple pairwise binary classification problem that classifies a pair of examples as similar or dissimilar.

28, TITLE:      A Framework for the Computational Linguistic Analysis of Dehumanization
http://arxiv.org/abs/2003.03014
AUTHORS:        Julia Mendelsohn ; Yulia Tsvetkov ; Dan Jurafsky
COMMENTS:       31 pages, 6 figures (Appendix is 4 pages, 4 figures). Submitted to Frontiers in Artificial Intelligence (Language and Computation)
HIGHLIGHT:      Drawing upon social psychology research, we create a computational linguistic framework for analyzing dehumanizing language by identifying linguistic correlates of salient components of dehumanization.

29, TITLE:      Exploring the ability of CNNs to generalise to previously unseen scales over wide scale ranges
http://arxiv.org/abs/2004.01536
AUTHORS:        Ylva Jansson ; Tony Lindeberg
COMMENTS:       14 pages, 6 figures
HIGHLIGHT:      We, therefore, present a theoretical analysis of invariance and covariance properties of scale channel networks and perform an experimental evaluation of the ability of different types of scale channel networks to generalise to previously unseen scales.

30, TITLE:      Artificial Intelligence: A Child's Play
http://arxiv.org/abs/1907.04659
AUTHORS:        Ravi Kashyap
HIGHLIGHT:      We discuss the objectives of any endeavor in creating artificial intelligence, AI, and provide a possible alternative.

31, TITLE:      Retrospective Analysis of the 2019 MineRL Competition on Sample Efficient Reinforcement Learning
http://arxiv.org/abs/2003.05012
AUTHORS:        Stephanie Milani ; Nicholay Topin ; Brandon Houghton ; William H. Guss ; Sharada P. Mohanty ; Keisuke Nakata ; Oriol Vinyals ; Noboru Sean Kuno
COMMENTS:       To appear in Proceedings of Machine Learning Research: NeurIPS 2019 Competition & Demonstration Track Postproceedings. 12 pages, 2 figures
HIGHLIGHT:      We describe the competition, outlining the primary challenge, the competition design, and the resources that we provided to the participants.

32, TITLE:      Adaptive Geo-Topological Independence Criterion
http://arxiv.org/abs/1810.02923
AUTHORS:        Baihan Lin ; Nikolaus Kriegeskorte
HIGHLIGHT:      We propose a class of adaptive (multi-threshold) test statistics, which form the basis for permutation tests.

33, 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, fix table formatting
HIGHLIGHT:      We propose a computationally lightweight method for approximating the variance matrix during training.

34, TITLE:      Semantic-driven Colorization
http://arxiv.org/abs/2006.07587
AUTHORS:        Man M. Ho ; Lu Zhang ; Alexander Raake ; Jinjia Zhou
COMMENTS:       This work is available at https://minhmanho.github.io/semantic-driven_colorization/
HIGHLIGHT:      In this study, we simulate that human-like action to firstly let our network learn to segment what is in the photo, then colorize it.

35, TITLE:      Robust parametric modeling of Alzheimer's disease progression
http://arxiv.org/abs/1908.05338
AUTHORS:        Mostafa Mehdipour Ghazi ; Mads Nielsen ; Akshay Pai ; Marc Modat ; M. Jorge Cardoso ; Sébastien Ourselin ; Lauge Sørensen
HIGHLIGHT:      Different M-estimators and logistic functions, including a novel type proposed in this study, called modified Stannard, are evaluated on the data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) for robust modeling of volumetric MRI and PET biomarkers, CSF measurements, as well as cognitive tests.

36, TITLE:      Frequency Fitness Assignment: Making Optimization Algorithms Invariant under Bijective Transformations of the Objective Function Value
http://arxiv.org/abs/2001.01416
AUTHORS:        Thomas Weise ; Zhize Wu ; Xinlu Li ; Yan Chen
HIGHLIGHT:      We verify this by applying the Md5 checksum computation as transformation to some of the above problems and yield the same behaviors.

37, TITLE:      Discovering Latent Classes for Semi-Supervised Semantic Segmentation
http://arxiv.org/abs/1912.12936
AUTHORS:        Olga Zatsarynna ; Johann Sawatzky ; Juergen Gall
HIGHLIGHT:      In order to leverage the information present in the unlabeled images, we propose to learn a second task that is related to semantic segmentation but easier.

38, TITLE:      Learning Dynamic Belief Graphs to Generalize on Text-Based Games
http://arxiv.org/abs/2002.09127
AUTHORS:        Ashutosh Adhikari ; Xingdi Yuan ; Marc-Alexandre Côté ; Mikuláš Zelinka ; Marc-Antoine Rondeau ; Romain Laroche ; Pascal Poupart ; Jian Tang ; Adam Trischler ; William L. Hamilton
HIGHLIGHT:      In this work, we investigate how an agent can plan and generalize in text-based games using graph-structured representations learned end-to-end from raw text.

39, TITLE:      Project CLAI: Instrumenting the Command Line as a New Environment for AI Agents
http://arxiv.org/abs/2002.00762
AUTHORS:        Mayank Agarwal ; Jorge J. Barroso ; Tathagata Chakraborti ; Eli M. Dow ; Kshitij Fadnis ; Borja Godoy ; Madhavan Pallan ; Kartik Talamadupula
COMMENTS:       http://ibm.biz/clai-home
HIGHLIGHT:      In this paper, we discuss the design and implementation of the platform in detail, through illustrative use cases of new end user interaction patterns enabled by this design, and through quantitative evaluation of the system footprint of a CLAI-enabled terminal.

40, TITLE:      Towards Automatic Embryo Staging in 3D+T Microscopy Images using Convolutional Neural Networks and PointNets
http://arxiv.org/abs/1910.00443
AUTHORS:        Manuel Traub ; Johannes Stegmaier
COMMENTS:       10 pages, 3 figures, 1 table
HIGHLIGHT:      In this contribution, we assess multiple approaches to perform automatic staging of developing embryos that were imaged with time-resolved 3D light-sheet microscopy.

41, TITLE:      An Algorithm for Computing Invariant Projectors in Representations of Wreath Products
http://arxiv.org/abs/1906.00858
AUTHORS:        Vladimir V. Kornyak
COMMENTS:       13 pages, version 2: computer outputs corrected
HIGHLIGHT:      We describe an algorithm for computing the complete set of primitive orthogonal idempotents in the centralizer ring of the permutation representation of a wreath product.

42, TITLE:      Contextual Bandit with Adaptive Feature Extraction
http://arxiv.org/abs/1802.00981
AUTHORS:        Baihan Lin ; Djallel Bouneffouf ; Guillermo Cecchi ; Irina Rish
COMMENTS:       Published in IEEE ICDMW 2018
HIGHLIGHT:      We consider an online decision making setting known as contextual bandit problem, and propose an approach for improving contextual bandit performance by using an adaptive feature extraction (representation learning) based on online clustering.

43, TITLE:      Learning to Play No-Press Diplomacy with Best Response Policy Iteration
http://arxiv.org/abs/2006.04635
AUTHORS:        Thomas Anthony ; Tom Eccles ; Andrea Tacchetti ; János Kramár ; Ian Gemp ; Thomas C. Hudson ; Nicolas Porcel ; Marc Lanctot ; Julien Pérolat ; Richard Everett ; Satinder Singh ; Thore Graepel ; Yoram Bachrach
HIGHLIGHT:      We propose a simple yet effective approximate best response operator, designed to handle large combinatorial action spaces and simultaneous moves.

44, TITLE:      Grasping in the Wild:Learning 6DoF Closed-Loop Grasping from Low-Cost Demonstrations
http://arxiv.org/abs/1912.04344
AUTHORS:        Shuran Song ; Andy Zeng ; Johnny Lee ; Thomas Funkhouser
COMMENTS:       Project Webpage https://graspinwild.cs.columbia.edu/
HIGHLIGHT:      In this work, we propose a new low-cost hardware interface for collecting grasping demonstrations by people in diverse environments.

45, TITLE:      An Interval-Valued Utility Theory for Decision Making with Dempster-Shafer Belief Functions
http://arxiv.org/abs/1912.06594
AUTHORS:        Thierry Denoeux ; Prakash P. Shenoy
HIGHLIGHT:      The main goal of this paper is to describe an axiomatic utility theory for Dempster-Shafer belief function lotteries.

46, TITLE:      Revisiting Training Strategies and Generalization Performance in Deep Metric Learning
http://arxiv.org/abs/2002.08473
AUTHORS:        Karsten Roth ; Timo Milbich ; Samarth Sinha ; Prateek Gupta ; Björn Ommer ; Joseph Paul Cohen
COMMENTS:       ICML 2020. Main paper 8 pages, 25 pages total
HIGHLIGHT:      Exploiting these insights, we propose a simple, yet effective, training regularization to reliably boost the performance of ranking-based DML models on various standard benchmark datasets.

47, 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
COMMENTS:       Submitted to Pattern Recognition
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.

48, TITLE:      A Comprehensive Review for Breast Histopathology Image Analysis Using Classical and Deep Neural Networks
http://arxiv.org/abs/2003.12255
AUTHORS:        Xiaomin Zhou ; Chen Li ; Md Mamunur Rahaman ; Yudong Yao ; Shiliang Ai ; Changhao Sun ; Xiaoyan Li ; Qian Wang ; Tao Jiang
COMMENTS:       25 pages,19 figures
HIGHLIGHT:      In this review, we present a comprehensive overview of the BHIA techniques based on ANNs.

49, TITLE:      A Monolingual Approach to Contextualized Word Embeddings for Mid-Resource Languages
http://arxiv.org/abs/2006.06202
AUTHORS:        Pedro Javier Ortiz Suárez ; Laurent Romary ; Benoît Sagot
HIGHLIGHT:      We use the multilingual OSCAR corpus, extracted from Common Crawl via language classification, filtering and cleaning, to train monolingual contextualized word embeddings (ELMo) for five mid-resource languages.

50, TITLE:      FUSE: Multi-Faceted Set Expansion by Coherent Clustering of Skip-grams
http://arxiv.org/abs/1910.04345
AUTHORS:        Wanzheng Zhu ; Hongyu Gong ; Jiaming Shen ; Chao Zhang ; Jingbo Shang ; Suma Bhat ; Jiawei Han
HIGHLIGHT:      In this paper, we study the task of multi-faceted set expansion, which aims to capture all semantic facets in the seed set and return multiple sets of entities, one for each semantic facet.

51, TITLE:      Proving P!=NP in first-order PA
http://arxiv.org/abs/2005.10080
AUTHORS:        Rupert McCallum
HIGHLIGHT:      We show that it is provable in PA that there is an arithmetically definable sequence $\{\phi_{n}:n \in \omega\}$ of $\Pi^{0}_{2}$-sentences, such that - PRA+$\{\phi_{n}:n \in \omega\}$ is $\Pi^{0}_{2}$-sound and $\Pi^{0}_{1}$-complete - the length of $\phi_{n}$ is bounded above by a polynomial function of $n$ with positive leading coefficient - PRA+$\phi_{n+1}$ always proves 1-consistency of PRA+$\phi_{n}$.

52, TITLE:      A Model of Fast Concept Inference with Object-Factorized Cognitive Programs
http://arxiv.org/abs/2002.04021
AUTHORS:        Daniel P. Sawyer ; Miguel Lázaro-Gredilla ; Dileep George
COMMENTS:       7 pages, 4 figures, 5 tables, to be presented at CogSci 2020
HIGHLIGHT:      To circumvent this bottleneck, we present an algorithm that emulates the human cognitive heuristics of object factorization and sub-goaling, allowing human-level inference speed, improving accuracy, and making the output more explainable.

53, TITLE:      Cross-view Semantic Segmentation for Sensing Surroundings
http://arxiv.org/abs/1906.03560
AUTHORS:        Bowen Pan ; Jiankai Sun ; Ho Yin Tiga Leung ; Alex Andonian ; Bolei Zhou
HIGHLIGHT:      To facilitate the robot perception with such a surrounding sensing capability, we introduce a novel visual task called Cross-view Semantic Segmentation as well as a framework named View Parsing Network (VPN) to address it.

54, 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.

55, TITLE:      Approximating Stacked and Bidirectional Recurrent Architectures with the Delayed Recurrent Neural Network
http://arxiv.org/abs/1909.00021
AUTHORS:        Javier S. Turek ; Shailee Jain ; Vy Vo ; Mihai Capota ; Alexander G. Huth ; Theodore L. Willke
COMMENTS:       to be published in Proceedings of International Conference on Machine Learning 2020 (ICML)
HIGHLIGHT:      In this work, we explore the delayed-RNN, which is a single-layer RNN that has a delay between the input and output.