2020.02.25.txt

==========New Papers==========
1, TITLE:       Extended formulation and valid inequalities for the multi-item inventory lot-sizing problem with supplier selection
http://arxiv.org/abs/2002.09916
AUTHORS:        Leopoldo E. Cárdenas-Barrón ; Rafael A. Melo ; Marcio C. Santos
HIGHLIGHT:      We propose a facility location extended formulation for the problem which can be preprocessed based on the cost structure and describe new valid inequalities in the original space of variables, which we denote $(l,S_j)$-inequalities.

2, TITLE:       Symbolic Learning and Reasoning with Noisy Data for Probabilistic Anchoring
http://arxiv.org/abs/2002.10373
AUTHORS:        Pedro Zuidberg Dos Martires ; Nitesh Kumar ; Andreas Persson ; Amy Loutfi ; Luc De Raedt
HIGHLIGHT:      We propose a semantic world modeling approach based on bottom-up object anchoring using an object-centered representation of the world.

3, TITLE:       Cognitive Argumentation and the Suppression Task
http://arxiv.org/abs/2002.10149
AUTHORS:        Emmanuelle-Anna Dietz Saldanha ; Antonis Kakas
HIGHLIGHT:      This paper addresses the challenge of modeling human reasoning, within a new framework called Cognitive Argumentation.

4, TITLE:       Optimizing Traffic Lights with Multi-agent Deep Reinforcement Learning and V2X communication
http://arxiv.org/abs/2002.09853
AUTHORS:        Azhar Hussain ; Tong Wang ; Cao Jiahua
COMMENTS:       7 Figure, Table 1
HIGHLIGHT:      We consider a system to optimize duration of traffic signals using multi-agent deep reinforcement learning and Vehicle-to-Everything (V2X) communication.

5, TITLE:       Injective Domain Knowledge in Neural Networks for Transprecision Computing
http://arxiv.org/abs/2002.10214
AUTHORS:        Andrea Borghesi ; Federico Baldo ; Michele Lombardi ; Michela Milano
HIGHLIGHT:      Fortunately, in many contexts domain knowledge is explicitly available and can be used to train better ML models.

6, TITLE:       KBSET -- Knowledge-Based Support for Scholarly Editing and Text Processing with Declarative LaTeX Markup and a Core Written in SWI-Prolog
http://arxiv.org/abs/2002.10329
AUTHORS:        Jana Kittelmann ; Christoph Wernhard
COMMENTS:       To appear in DECLARE 2019 Revised Selected Papers
HIGHLIGHT:      KBSET -- Knowledge-Based Support for Scholarly Editing and Text Processing with Declarative LaTeX Markup and a Core Written in SWI-Prolog

7, TITLE:       Signature in Counterparts, a Formal Treatment
http://arxiv.org/abs/2002.09827
AUTHORS:        Ron van der Meyden
HIGHLIGHT:      The paper develops a logical understanding of this process, developing a number of axioms that can be used to justify the validity of a contract from the assumption that separate copies have been signed.

8, TITLE:       Symbolic Querying of Vector Spaces: Probabilistic Databases Meets Relational Embeddings
http://arxiv.org/abs/2002.10029
AUTHORS:        Tal Friedman ; Guy Van den Broeck
HIGHLIGHT:      To deal with increasing amounts of uncertainty and incompleteness in relational data, we propose unifying techniques from probabilistic databases and relational embedding models.

9, TITLE:       Markov Logic Networks with Complex Weights: Expressivity, Liftability and Fourier Transforms
http://arxiv.org/abs/2002.10259
AUTHORS:        Ondrej Kuzelka
HIGHLIGHT:      We introduce complex MLNs, which use complex-valued weights, and we show that, unlike standard MLNs with real-valued weights, complex MLNs are fully expressive.

10, TITLE:      Hyperbolic Minesweeper is in P
http://arxiv.org/abs/2002.09534
AUTHORS:        Eryk Kopczyński
HIGHLIGHT:      We show that, while Minesweeper is NP-complete, its hyperbolic variant is in P.

11, TITLE:      Triple Wins: Boosting Accuracy, Robustness and Efficiency Together by Enabling Input-Adaptive Inference
http://arxiv.org/abs/2002.10025
AUTHORS:        Ting-Kuei Hu ; Tianlong Chen ; Haotao Wang ; Zhangyang Wang
HIGHLIGHT:      That multi-loss adaptivity adds new variations and flexibility to adversarial attacks and defenses, on which we present a systematical investigation.

12, TITLE:      DotFAN: A Domain-transferred Face Augmentation Network for Pose and Illumination Invariant Face Recognition
http://arxiv.org/abs/2002.09859
AUTHORS:        Hao-Chiang Shao ; Kang-Yu Liu ; Chia-Wen Lin ; Jiwen Lu
COMMENTS:       12 pages, 10 figures
HIGHLIGHT:      In this paper, we propose a 3D model-assisted domain-transferred face augmentation network (DotFAN) that can generate a series of variants of an input face based on the knowledge distilled from existing rich face datasets collected from other domains.

13, TITLE:      Multi-Stream Networks and Ground-Truth Generation for Crowd Counting
http://arxiv.org/abs/2002.09951
AUTHORS:        Rodolfo Quispe ; Darwin Ttito ; Adín Rivera ; Helio Pedrini
COMMENTS:       https://github.com/RQuispeC/multi-stream-crowd-counting-extended
HIGHLIGHT:      A Multi-Stream Convolutional Neural Network is developed and evaluated in this work, which receives an image as input and produces a density map that represents the spatial distribution of people in an end-to-end fashion.

14, TITLE:      Exploring Spatial-Temporal Multi-Frequency Analysis for High-Fidelity and Temporal-Consistency Video Prediction
http://arxiv.org/abs/2002.09905
AUTHORS:        Beibei Jin ; Yu Hu ; Qiankun Tang ; Jingyu Niu ; Zhiping Shi ; Yinhe Han ; Xiaowei Li
COMMENTS:       Submitted to a conference, under review
HIGHLIGHT:      In this paper, we point out the necessity of exploring multi-frequency analysis to deal with the two problems.

15, TITLE:      Assembling Semantically-Disentangled Representations for Predictive-Generative Models via Adaptation from Synthetic Domain
http://arxiv.org/abs/2002.09818
AUTHORS:        Burkay Donderici ; Caleb New ; Chenliang Xu
COMMENTS:       8 pages, 18 figures
HIGHLIGHT:      In this paper, we show that semantically-aligned representations can be generated instead with the help of a physics based engine.

16, TITLE:      Random Bundle: Brain Metastases Segmentation Ensembling through Annotation Randomization
http://arxiv.org/abs/2002.09809
AUTHORS:        Darvin Yi ; Endre Gøvik ; Michael Iv ; Elizabeth Tong ; Greg Zaharchuk ; Daniel Rubin
HIGHLIGHT:      We introduce a novel ensembling method, Random Bundle (RB), that improves performance for brain metastases segmentation.

17, TITLE:      Deep Multimodal Image-Text Embeddings for Automatic Cross-Media Retrieval
http://arxiv.org/abs/2002.10016
AUTHORS:        Hadi Abdi Khojasteh ; Ebrahim Ansari ; Parvin Razzaghi ; Akbar Karimi
COMMENTS:       6 pages and 2 figures, Learn more about this project at https://iasbs.ac.ir/~ansari/deeptwitter
HIGHLIGHT:      In this work, we introduce an end-to-end deep multimodal convolutional-recurrent network for learning both vision and language representations simultaneously to infer image-text similarity.

18, TITLE:      FONDUE: A Framework for Node Disambiguation Using Network Embeddings
http://arxiv.org/abs/2002.10127
AUTHORS:        Ahmad Mel ; Bo Kang ; Jefrey Lijffijt ; Tijl De Bie
COMMENTS:       11 pages, 3 figures
HIGHLIGHT:      This can be valuable for a range of purposes from the study of information diffusion to bibliographic analysis, bioinformatics research, and question-answering.

19, TITLE:      Practical and Bilateral Privacy-preserving Federated Learning
http://arxiv.org/abs/2002.09843
AUTHORS:        Yan Feng ; Xue Yang ; Weijun Fang ; Shu-Tao Xia ; Xiaohu Tang
COMMENTS:       Submitted to ICML 2020
HIGHLIGHT:      In this paper, we present the first bilateral privacy-preserving federated learning scheme, which protects not only the raw training data of clients, but also model iterates during the training phase as well as final model parameters.

20, TITLE:      Deep Reinforcement Learning with Linear Quadratic Regulator Regions
http://arxiv.org/abs/2002.09820
AUTHORS:        Gabriel I. Fernandez ; Colin Togashi ; Dennis Hong ; Lin Yang
HIGHLIGHT:      In this paper we propose a novel method that guarantees a stable region of attraction for the output of a policy trained in simulation, even for highly nonlinear systems.

21, TITLE:      Variance Loss in Variational Autoencoders
http://arxiv.org/abs/2002.09860
AUTHORS:        Andrea Asperti
HIGHLIGHT:      In this article, we highlight what appears to be major issue of Variational Autoencoders, evinced from an extensive experimentation with different network architectures and datasets: the variance of generated data is sensibly lower than that of training data.

22, TITLE:      Structural Combinatorial of Network Information System of Systems based on Evolutionary Optimization Method
http://arxiv.org/abs/2002.09706
AUTHORS:        Tingting Zhang ; Yushi Lan ; Aiguo Song ; Kun Liu ; Nan Wang
HIGHLIGHT:      Given that the concept of evolution originates from biological systems, in this article, the evolution of network information architecture is analyzed by genetic algorithms, and the network information architecture is represented by chromosomes.

23, TITLE:      Superoptimization of WebAssembly Bytecode
http://arxiv.org/abs/2002.10213
AUTHORS:        Javier Cabrera Arteaga ; Shrinish Donde ; Jian Gu ; Orestis Floros ; Lucas Satabin ; Benoit Baudry ; Martin Monperrus
COMMENTS:       4 pages, 3 figures, MoreVMs 2020
HIGHLIGHT:      Motivated by the fast adoption of WebAssembly, we propose the first functional pipeline to support the superoptimization of WebAssembly bytecode.

24, TITLE:      Beyond Camera Motion Removing: How to Handle Outliers in Deblurring
http://arxiv.org/abs/2002.10201
AUTHORS:        Chenwei Yang ; Meng Chang ; Huajun Feng ; Zhihai Xu ; Qi Li
HIGHLIGHT:      We propose a salient edge detection network to supervise the training process and solve the outlier problem by proposing a novel method of dataset generation.

25, TITLE:      Convex Shape Representation with Binary Labels for Image Segmentation: Models and Fast Algorithms
http://arxiv.org/abs/2002.09600
AUTHORS:        Shousheng Luo ; Xue-Cheng Tai ; Yang Wang
HIGHLIGHT:      We present a novel and effective binary representation for convex shapes.

26, TITLE:      Robust Numerical Tracking of One Path of a Polynomial Homotopy on Parallel Shared Memory Computers
http://arxiv.org/abs/2002.09504
AUTHORS:        Simon Telen ; Marc Van Barel ; Jan Verschelde
HIGHLIGHT:      We consider the problem of tracking one solution path defined by a polynomial homotopy on a parallel shared memory computer.

27, TITLE:      From Chess and Atari to StarCraft and Beyond: How Game AI is Driving the World of AI
http://arxiv.org/abs/2002.10433
AUTHORS:        Sebastian Risi ; Mike Preuss
HIGHLIGHT:      Because of the growth of the field, a single review cannot cover it completely.

28, TITLE:      "Wait, I'm Still Talking!" Predicting the Dialogue Interaction Behavior Using Imagine-Then-Arbitrate Model
http://arxiv.org/abs/2002.09616
AUTHORS:        Zehao Lin ; Xiaoming Kang ; Guodun Li ; Feng Ji ; Haiqing Chen ; Yin Zhang
HIGHLIGHT:      To address this issue, in this paper, we propose a novel Imagine-then-Arbitrate (ITA) neural dialogue model to help the agent decide whether to wait or to make a response directly.

29, TITLE:      Data Augmentation for Copy-Mechanism in Dialogue State Tracking
http://arxiv.org/abs/2002.09634
AUTHORS:        Xiaohui Song ; Liangjun Zang ; Yipeng Su ; Xing Wu ; Jizhong Han ; Songlin Hu
HIGHLIGHT:      In this paper, we aim to find out the factors that influence the generalization ability of a common copy-mechanism model for DST.

30, TITLE:      Incorporating Effective Global Information via Adaptive Gate Attention for Text Classification
http://arxiv.org/abs/2002.09673
AUTHORS:        Xianming Li ; Zongxi Li ; Yingbin Zhao ; Haoran Xie ; Qing Li
HIGHLIGHT:      In this paper, we propose a classifier with gate mechanism named Adaptive Gate Attention model with Global Information (AGA+GI), in which the adaptive gate mechanism incorporates global statistical features into latent semantic features and the attention layer captures dependency relationship within the sentence.

31, TITLE:      Markov Chain Monte-Carlo Phylogenetic Inference Construction in Computational Historical Linguistics
http://arxiv.org/abs/2002.09637
AUTHORS:        Tianyi Ni
HIGHLIGHT:      In this paper, I am going to use computational method to cluster the languages and use Markov Chain Monte Carlo (MCMC) method to build the language typology relationship tree based on the clusters.

32, TITLE:      Emergent Communication with World Models
http://arxiv.org/abs/2002.09604
AUTHORS:        Alexander I. Cowen-Rivers ; Jason Naradowsky
COMMENTS:       NeurIPS Workshop on Emergent Communication
HIGHLIGHT:      We introduce Language World Models, a class of language-conditional generative model which interpret natural language messages by predicting latent codes of future observations.

33, TITLE:      Efficient Sentence Embedding via Semantic Subspace Analysis
http://arxiv.org/abs/2002.09620
AUTHORS:        Bin Wang ; Fenxiao Chen ; Yuncheng Wang ; C. -C. Jay Kuo
COMMENTS:       7 pages, 2 figures
HIGHLIGHT:      A novel sentence embedding method built upon semantic subspace analysis, called semantic subspace sentence embedding (S3E), is proposed in this work.

34, TITLE:      Machine Translation System Selection from Bandit Feedback
http://arxiv.org/abs/2002.09646
AUTHORS:        Jason Naradowsky ; Xuan Zhang ; Kevin Duh
HIGHLIGHT:      In this work we take a different approach, treating the problem of adapting as one of selection.

35, TITLE:      Exploiting Typed Syntactic Dependencies for Targeted Sentiment Classification Using Graph Attention Neural Network
http://arxiv.org/abs/2002.09685
AUTHORS:        Xuefeng Bai ; Pengbo Liu ; Yue Zhang
HIGHLIGHT:      Dominant methods employ neural networks for encoding the input sentence and extracting relations between target mentions and their contexts.

36, TITLE:      HRank: Filter Pruning using High-Rank Feature Map
http://arxiv.org/abs/2002.10179
AUTHORS:        Mingbao Lin ; Rongrong Ji ; Yan Wang ; Yichen Zhang ; Baochang Zhang ; Yonghong Tian ; Ling Shao
HIGHLIGHT:      In this paper, we propose a novel filter pruning method by exploring the High Rank of feature maps (HRank).

37, TITLE:      When Relation Networks meet GANs: Relation GANs with Triplet Loss
http://arxiv.org/abs/2002.10174
AUTHORS:        Runmin Wu ; Kunyao Zhang ; Lijun Wang ; Yue Wang ; Huchuan Lu ; Yizhou Yu
COMMENTS:       8 pages
HIGHLIGHT:      In this paper, we provide a more principled approach as an alternative solution to this issue.

38, TITLE:      Improving STDP-based Visual Feature Learning with Whitening
http://arxiv.org/abs/2002.10177
AUTHORS:        Pierre Falez ; Pierre Tirilly ; Ioan Marius Bilasco
HIGHLIGHT:      In this paper, we propose to use whitening as a pre-processing step before learning features with STDP.

39, TITLE:      Audio-driven Talking Face Video Generation with Natural Head Pose
http://arxiv.org/abs/2002.10137
AUTHORS:        Ran Yi ; Zipeng Ye ; Juyong Zhang ; Hujun Bao ; Yong-Jin Liu
COMMENTS:       12 pages, 7 figures
HIGHLIGHT:      In this paper, we address this problem by proposing a deep neural network model that takes an audio signal A of a source person and a very short video V of a target person as input, and outputs a synthesized high-quality talking face video with natural head pose (making use of the visual information in V), expression and lip synchronization (by considering both A and V).

40, TITLE:      3DSSD: Point-based 3D Single Stage Object Detector
http://arxiv.org/abs/2002.10187
AUTHORS:        Zetong Yang ; Yanan Sun ; Shu Liu ; Jiaya Jia
HIGHLIGHT:      We novelly propose a fusion sampling strategy in downsampling process to make detection on less representative points feasible.

41, TITLE:      Emosaic: Visualizing Affective Content of Text at Varying Granularity
http://arxiv.org/abs/2002.10096
AUTHORS:        Philipp Geuder ; Marie Claire Leidinger ; Martin von Lupin ; Marian Dörk ; Tobias Schröder
COMMENTS:       9 pages, 7 figures
HIGHLIGHT:      This paper presents Emosaic, a tool for visualizing the emotional tone of text documents, considering multiple dimensions of emotion and varying levels of semantic granularity.

42, TITLE:      Leveraging Code Generation to Improve Code Retrieval and Summarization via Dual Learning
http://arxiv.org/abs/2002.10198
AUTHORS:        Wei Ye ; Rui Xie ; Jinglei Zhang ; Tianxiang Hu ; Xiaoyin Wang ; Shikun Zhang
COMMENTS:       Published at The Web Conference (WWW) 2020, full paper
HIGHLIGHT:      In this paper, we propose a novel end-to-end model for the two tasks by introducing an additional code generation task.

43, TITLE:      ORCSolver: An Efficient Solver for Adaptive GUI Layout with OR-Constraints
http://arxiv.org/abs/2002.09925
AUTHORS:        Yue Jiang ; Wolfgang Stuerzlinger ; Matthias Zwicker ; Christof Lutteroth
COMMENTS:       Published at CHI2020
HIGHLIGHT:      To address this challenge, we propose ORCSolver, a novel solving technique for adaptive ORC layouts, based on a branch-and-bound approach with heuristic preprocessing.

44, TITLE:      The Early Phase of Neural Network Training
http://arxiv.org/abs/2002.10365
AUTHORS:        Jonathan Frankle ; David J. Schwab ; Ari S. Morcos
COMMENTS:       ICLR 2020 Camera Ready. Available on OpenReview at https://openreview.net/forum?id=Hkl1iRNFwS
HIGHLIGHT:      We perform extensive measurements of the network state during these early iterations of training and leverage the framework of Frankle et al. (2019) to quantitatively probe the weight distribution and its reliance on various aspects of the dataset.

45, TITLE:      Self-Adaptive Training: beyond Empirical Risk Minimization
http://arxiv.org/abs/2002.10319
AUTHORS:        Lang Huang ; Chao Zhang ; Hongyang Zhang
COMMENTS:       13 pages, 7 figures, 5 tables
HIGHLIGHT:      In this paper, we observe that model predictions can substantially benefit the training process: self-adaptive training significantly improves generalization over ERM under various levels of noises, and mitigates the overfitting issue in both natural and adversarial training.

46, TITLE:      Gradual Channel Pruning while Training using Feature Relevance Scores for Convolutional Neural Networks
http://arxiv.org/abs/2002.09958
AUTHORS:        Sai Aparna Aketi ; Sourjya Roy ; Anand Raghunathan ; Kaushik Roy
COMMENTS:       10 pages, 6 figures, 5 tables
HIGHLIGHT:      To address all the above issues, we present a simple-yet-effective gradual channel pruning while training methodology using a novel data driven metric referred as Feature relevance score.

47, TITLE:      Temporal Spike Sequence Learning via Backpropagation for Deep Spiking Neural Networks
http://arxiv.org/abs/2002.10085
AUTHORS:        Wenrui Zhang ; Peng Li
HIGHLIGHT:      We present a novel Temporal Spike Sequence Learning Backpropagation (TSSL-BP) method for training deep SNNs, which breaks down error backpropagation across two types of inter-neuron and intra-neuron dependencies.

48, TITLE:      Towards Label-Free 3D Segmentation of Optical Coherence Tomography Images of the Optic Nerve Head Using Deep Learning
http://arxiv.org/abs/2002.09635
AUTHORS:        Sripad Krishna Devalla ; Tan Hung Pham ; Satish Kumar Panda ; Liang Zhang ; Giridhar Subramanian ; Anirudh Swaminathan ; Chin Zhi Yun ; Mohan Rajan ; Sujatha Mohan ; Ramaswami Krishnadas ; Vijayalakshmi Senthil ; John Mark S. de Leon ; Tin A. Tun ; Ching-Yu Cheng ; Leopold Schmetterer ; Shamira Perera ; Tin Aung ; Alexandre H. Thiery ; Micha el J. A. Girard
HIGHLIGHT:      To address this, we propose a DL based 3D segmentation framework that is easily translatable across OCT devices in a label-free manner (i.e. without the need to manually re-segment data for each device). Specifically, we developed 2 sets of DL networks.

49, TITLE:      Unsupervised Denoising for Satellite Imagery using Wavelet Subband CycleGAN
http://arxiv.org/abs/2002.09847
AUTHORS:        Joonyoung Song ; Jae-Heon Jeong ; Dae-Soon Park ; Hyun-Ho Kim ; Doo-Chun Seo ; Jong Chul Ye
HIGHLIGHT:      In this paper, we propose a novel unsupervised multispectral denoising method for satellite imagery using wavelet subband cycle-consistent adversarial network (WavCycleGAN). Moreover, in contrast to the standard image domain cycleGAN, we introduce a wavelet subband domain learning scheme for effective denoising without sacrificing high frequency components such as edges and detail information.

50, TITLE:      Tuning-free Plug-and-Play Proximal Algorithm for Inverse Imaging Problems
http://arxiv.org/abs/2002.09611
AUTHORS:        Kaixuan Wei ; Angelica Aviles-Rivero ; Jingwei Liang ; Ying Fu ; Carola-Bibiane Schnlieb ; Hua Huang
HIGHLIGHT:      In this work, we present a tuning-free PnP proximal algorithm, which can automatically determine the internal parameters including the penalty parameter, the denoising strength and the terminal time.

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

52, TITLE:      Generalized Octave Convolutions for Learned Multi-Frequency Image Compression
http://arxiv.org/abs/2002.10032
AUTHORS:        Mohammad Akbari ; Jie Liang ; Jingning Han ; Chengjie Tu
COMMENTS:       10 pages, 7 figures, 3 tables
HIGHLIGHT:      In this paper, we propose the first learned multi-frequency image compression approach that uses the recently developed octave convolutions to factorize the latents into high and low frequencies.

53, TITLE:      Automatic Data Augmentation via Deep Reinforcement Learning for Effective Kidney Tumor Segmentation
http://arxiv.org/abs/2002.09703
AUTHORS:        Tiexin Qin ; Ziyuan Wang ; Kelei He ; Yinghuan Shi ; Yang Gao ; Dinggang Shen
COMMENTS:       5 pages, 3 figures
HIGHLIGHT:      In this paper, we developed a novel automatic learning-based data augmentation method for medical image segmentation which models the augmentation task as a trial-and-error procedure using deep reinforcement learning (DRL).

54, TITLE:      Computer-inspired Quantum Experiments
http://arxiv.org/abs/2002.09970
AUTHORS:        Mario Krenn ; Manuel Erhard ; Anton Zeilinger
COMMENTS:       Comments and suggestions for additional references are welcome!
HIGHLIGHT:      We focus on the most mature and \textit{practical} approaches that scientists used to find new complex quantum experiments, which experimentalists subsequently have realized in the laboratories.

55, TITLE:      A Nepali Rule Based Stemmer and its performance on different NLP applications
http://arxiv.org/abs/2002.09901
AUTHORS:        Pravesh Koirala ; Aman Shakya
COMMENTS:       5 pages, 2 figures, 3 tables
HIGHLIGHT:      This study focuses on creating a Rule Based stemmer for Nepali text.

56, TITLE:      Do Multi-Hop Question Answering Systems Know How to Answer the Single-Hop Sub-Questions?
http://arxiv.org/abs/2002.09919
AUTHORS:        Yixuan Tang ; Hwee Tou Ng ; Anthony K. H. Tung
HIGHLIGHT:      In this paper, we investigate whether top-performing models for multi-hop questions understand the underlying sub-questions like humans.

57, TITLE:      Unsupervised Question Decomposition for Question Answering
http://arxiv.org/abs/2002.09758
AUTHORS:        Ethan Perez ; Patrick Lewis ; Wen-tau Yih ; Kyunghyun Cho ; Douwe Kiela
HIGHLIGHT:      Since collecting labeled decompositions is cumbersome, we propose an unsupervised approach to produce sub-questions.

58, TITLE:      Fill in the BLANC: Human-free quality estimation of document summaries
http://arxiv.org/abs/2002.09836
AUTHORS:        Oleg Vasilyev ; Vedant Dharnidharka ; John Bohannon
COMMENTS:       12 pages, 9 figures, 2 tables
HIGHLIGHT:      We present BLANC, a new approach to the automatic estimation of document summary quality.

59, TITLE:      PUGeo-Net: A Geometry-centric Network for 3D Point Cloud Upsampling
http://arxiv.org/abs/2002.10277
AUTHORS:        Yue Qian ; Junhui Hou ; Sam Kwong ; Ying He
HIGHLIGHT:      To tackle the challenge, we propose a novel deep neural network based method, called PUGeo-Net, that learns a $3\times 3$ linear transformation matrix $\bf T$ for each input point.

60, TITLE:      ABCNet: Real-time Scene Text Spotting with Adaptive Bezier-Curve Network
http://arxiv.org/abs/2002.10200
AUTHORS:        Yuliang Liu ; Hao Chen ; Chunhua Shen ; Tong He ; Lianwen Jin ; Liangwei Wang
COMMENTS:       Accepted to Proc. IEEE Conf. Comp. Vis. Pattern Recogn. (CVPR) 2020
HIGHLIGHT:      Our contributions are three-fold: 1) For the first time, we adaptively fit arbitrarily-shaped text by a parameterized Bezier curve.

61, TITLE:      Learning Attentive Pairwise Interaction for Fine-Grained Classification
http://arxiv.org/abs/2002.10191
AUTHORS:        Peiqin Zhuang ; Yali Wang ; Yu Qiao
COMMENTS:       Accepted at AAAI-2020
HIGHLIGHT:      Inspired by this fact, this paper proposes a simple but effective Attentive Pairwise Interaction Network (API-Net), which can progressively recognize a pair of fine-grained images by interaction.

62, TITLE:      On the General Value of Evidence, and Bilingual Scene-Text Visual Question Answering
http://arxiv.org/abs/2002.10215
AUTHORS:        Xinyu Wang ; Yuliang Liu ; Chunhua Shen ; Chun Chet Ng ; Canjie Luo ; Lianwen Jin ; Chee Seng Chan ; Anton van den Hengel ; Liangwei Wang
COMMENTS:       Accepted to Proc. IEEE Conf. Computer Vision and Pattern Recognition 2020
HIGHLIGHT:      We present a dataset that takes a step towards addressing this problem in that it contains questions expressed in two languages, and an evaluation process that co-opts a well understood image-based metric to reflect the method's ability to reason.

63, TITLE:      Mnemonics Training: Multi-Class Incremental Learning without Forgetting
http://arxiv.org/abs/2002.10211
AUTHORS:        Yaoyao Liu ; An-An Liu ; Yuting Su ; Bernt Schiele ; Qianru Sun
COMMENTS:       Accepted by CVPR 2020
HIGHLIGHT:      This paper proposes a novel and automatic framework we call mnemonics, where we parameterize exemplars and make them optimizable in an end-to-end manner.

64, TITLE:      Automatic Estimation of Sphere Centers from Images of Calibrated Cameras
http://arxiv.org/abs/2002.10217
AUTHORS:        Levente Hajder ; Tekla Tóth ; Zoltán Pusztai
HIGHLIGHT:      We propose two novel methods to (i) detect an ellipse in camera images and (ii) estimate the spatial location of the corresponding sphere if its size is known.

65, TITLE:      Image to Language Understanding: Captioning approach
http://arxiv.org/abs/2002.09536
AUTHORS:        Madhavan Seshadri ; Malavika Srikanth ; Mikhail Belov
COMMENTS:       8 pages
HIGHLIGHT:      This project aims to compare different approaches for solving the image captioning problem.

66, TITLE:      Group Membership Verification with Privacy: Sparse or Dense?
http://arxiv.org/abs/2002.10362
AUTHORS:        Marzieh Gheisari ; Teddy Furon ; Laurent Amsaleg
HIGHLIGHT:      This paper proposes a mathematical model for group membership verification allowing to reveal the impact of sparsity on both security, compactness, and verification performances.

67, TITLE:      Batch Normalization Biases Deep Residual Networks Towards Shallow Paths
http://arxiv.org/abs/2002.10444
AUTHORS:        Soham De ; Samuel L. Smith
HIGHLIGHT:      We identify the origin of this benefit: At initialization, batch normalization downscales the residual branch relative to the skip connection, by a normalizing factor proportional to the square root of the network depth.

68, TITLE:      Deep Nearest Neighbor Anomaly Detection
http://arxiv.org/abs/2002.10445
AUTHORS:        Liron Bergman ; Niv Cohen ; Yedid Hoshen
HIGHLIGHT:      In this work, we investigate whether the recent progress can indeed outperform nearest-neighbor methods operating on an Imagenet pretrained feature space.

69, TITLE:      Implicit Geometric Regularization for Learning Shapes
http://arxiv.org/abs/2002.10099
AUTHORS:        Amos Gropp ; Lior Yariv ; Niv Haim ; Matan Atzmon ; Yaron Lipman
HIGHLIGHT:      In this paper we offer a new paradigm for computing high fidelity implicit neural representations directly from raw data (i.e., point clouds, with or without normal information).

70, TITLE:      HarDNN: Feature Map Vulnerability Evaluation in CNNs
http://arxiv.org/abs/2002.09786
AUTHORS:        Abdulrahman Mahmoud ; Siva Kumar Sastry Hari ; Christopher W. Fletcher ; Sarita V. Adve ; Charbel Sakr ; Naresh Shanbhag ; Pavlo Molchanov ; Michael B. Sullivan ; Timothy Tsai ; Stephen W. Keckler
COMMENTS:       14 pages, 5 figures, a short version accepted for publication in First Workshop on Secure and Resilient Autonomy (SARA) co-located with MLSys2020
HIGHLIGHT:      This paper presents HarDNN, a software-directed approach to identify vulnerable computations during a CNN inference and selectively protect them based on their propensity towards corrupting the inference output in the presence of a hardware error.

71, TITLE:      Towards Robust and Reproducible Active Learning Using Neural Networks
http://arxiv.org/abs/2002.09564
AUTHORS:        Prateek Munjal ; Nasir Hayat ; Munawar Hayat ; Jamshid Sourati ; Shadab Khan
HIGHLIGHT:      In this study, we show that recent AL methods offer a gain over random baseline under a brittle combination of experimental conditions.

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

73, TITLE:      Towards precise causal effect estimation from data with hidden variables
http://arxiv.org/abs/2002.10091
AUTHORS:        Debo Cheng ; Jiuyong Li ; Lin Liu ; Kui Yu ; Thuc Duy Lee ; Jixue Liu
HIGHLIGHT:      In this paper, we identify a practical problem setting and propose an approach to achieving unique causal effect estimation from data with hidden variables under this setting.

74, TITLE:      Training Question Answering Models From Synthetic Data
http://arxiv.org/abs/2002.09599
AUTHORS:        Raul Puri ; Ryan Spring ; Mostofa Patwary ; Mohammad Shoeybi ; Bryan Catanzaro
HIGHLIGHT:      This work aims to narrow this gap by taking advantage of large language models and explores several factors such as model size, quality of pretrained models, scale of data synthesized, and algorithmic choices.

75, TITLE:      Modelling Latent Skills for Multitask Language Generation
http://arxiv.org/abs/2002.09543
AUTHORS:        Kris Cao ; Dani Yogatama
HIGHLIGHT:      We present a generative model for multitask conditional language generation.

76, TITLE:      Extracting and Validating Explanatory Word Archipelagoes using Dual Entropy
http://arxiv.org/abs/2002.09581
AUTHORS:        Yukio Ohsawa ; Teruaki Hayashi
COMMENTS:       7 pages, 2 figures, 2 columns
HIGHLIGHT:      Extracting and Validating Explanatory Word Archipelagoes using Dual Entropy

77, TITLE:      Hardness of equations over finite solvable groups under the exponential time hypothesis
http://arxiv.org/abs/2002.10145
AUTHORS:        Armin Weiß
HIGHLIGHT:      In this work, we present the first lower bounds for the equation satisfiability problem in finite solvable groups: under the assumption of the exponential time hypothesis, we show that it cannot be in P for any group of Fitting length at least four and for certain groups of Fitting length three.

78, TITLE:      Utilizing a null class to restrict decision spaces and defend against neural network adversarial attacks
http://arxiv.org/abs/2002.10084
AUTHORS:        Matthew J. Roos
COMMENTS:       15 pages, 19 figures
HIGHLIGHT:      In this paper we examine an additional, and largely unexplored, cause behind this phenomenon--namely, the use of the conventional training paradigm in which the entire input space is parcellated among the training classes.

79, TITLE:      DeepSign: Deep On-Line Signature Verification
http://arxiv.org/abs/2002.10119
AUTHORS:        Ruben Tolosana ; Ruben Vera-Rodriguez ; Julian Fierrez ; Javier Ortega-Garcia
HIGHLIGHT:      The main contributions of this study are: i) we provide an in-depth analysis of state-of-the-art deep learning approaches for on-line signature verification, ii) we present and describe the new DeepSignDB on-line handwritten signature biometric public database, iii) we propose a standard experimental protocol and benchmark to be used for the research community in order to perform a fair comparison of novel approaches with the state of the art, and iv) we adapt and evaluate our recent deep learning approach named Time-Aligned Recurrent Neural Networks (TA-RNNs) for the task of on-line handwritten signature verification.

80, TITLE:      SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint Estimation
http://arxiv.org/abs/2002.10111
AUTHORS:        Zechen Liu ; Zizhang Wu ; Roland Tóth
COMMENTS:       8 pages, 6 figures
HIGHLIGHT:      Hence, we propose a novel 3D object detection method, named SMOKE, in this paper that predicts a 3D bounding box for each detected object by combining a single keypoint estimate with regressed 3D variables.

81, TITLE:      LeafGAN: An Effective Data Augmentation Method for Practical Plant Disease Diagnosis
http://arxiv.org/abs/2002.10100
AUTHORS:        Quan Huu Cap ; Hiroyuki Uga ; Satoshi Kagiwada ; Hitoshi Iyatomi
HIGHLIGHT:      In this paper, we propose LeafGAN, a novel image-to-image translation system with own attention mechanism.

82, TITLE:      Semantic Flow for Fast and Accurate Scene Parsing
http://arxiv.org/abs/2002.10120
AUTHORS:        Xiangtai Li ; Ansheng You ; Zhen Zhu ; Houlong Zhao ; Maoke Yang ; Kuiyuan Yang ; Yunhai Tong
HIGHLIGHT:      In this paper, we focus on effective methods for fast and accurate scene parsing.

83, TITLE:      GANHopper: Multi-Hop GAN for Unsupervised Image-to-Image Translation
http://arxiv.org/abs/2002.10102
AUTHORS:        Wallace Lira ; Johannes Merz ; Daniel Ritchie ; Daniel Cohen-Or ; Hao Zhang
COMMENTS:       9 pages, 9 figures
HIGHLIGHT:      We introduce GANHOPPER, an unsupervised image-to-image translation network that transforms images gradually between two domains, through multiple hops.

84, TITLE:      Sketching Transformed Matrices with Applications to Natural Language Processing
http://arxiv.org/abs/2002.09812
AUTHORS:        Yingyu Liang ; Zhao Song ; Mengdi Wang ; Lin F. Yang ; Xin Yang
COMMENTS:       AISTATS 2020
HIGHLIGHT:      We show that our approach obtains small error and is efficient in both space and time.

85, TITLE:      Mixed Integer Programming for Searching Maximum Quasi-Bicliques
http://arxiv.org/abs/2002.09880
AUTHORS:        Dmitry I. Ignatov ; Polina Ivanova ; Albina Zamaletdinova
COMMENTS:       This paper draft is stored here for self-archiving purposes
HIGHLIGHT:      Several models based on Mixed Integer Programming (MIP) to search for a quasi-biclique are proposed and tested for working efficiency.

86, TITLE:      SupRB: A Supervised Rule-based Learning System for Continuous Problems
http://arxiv.org/abs/2002.10295
AUTHORS:        Michael Heider ; David Pätzel ; Jörg Hähner
COMMENTS:       Submitted to the Genetic and Evolutionary Computation Conference 2020 (GECCO 2020)
HIGHLIGHT:      We propose the SupRB learning system, a new Pittsburgh-style learning classifier system (LCS) for supervised learning on multi-dimensional continuous decision problems.

87, TITLE:      Learning Certified Individually Fair Representations
http://arxiv.org/abs/2002.10312
AUTHORS:        Anian Ruoss ; Mislav Balunović ; Marc Fischer ; Martin Vechev
HIGHLIGHT:      In this work, we introduce the first method which generalizes individual fairness to rich similarity notions via logical constraints while also enabling data consumers to obtain fairness certificates for their models.

88, TITLE:      Discriminative Particle Filter Reinforcement Learning for Complex Partial Observations
http://arxiv.org/abs/2002.09884
AUTHORS:        Xiao Ma ; Peter Karkus ; David Hsu ; Wee Sun Lee ; Nan Ye
COMMENTS:       Accepted to ICLR 2020
HIGHLIGHT:      This paper presents Discriminative Particle Filter Reinforcement Learning (DPFRL), a new reinforcement learning framework for complex partial observations.

89, TITLE:      Learning to Continually Learn
http://arxiv.org/abs/2002.09571
AUTHORS:        Shawn Beaulieu ; Lapo Frati ; Thomas Miconi ; Joel Lehman ; Kenneth O. Stanley ; Jeff Clune ; Nick Cheney
HIGHLIGHT:      Inspired by neuromodulatory processes in the brain, we propose A Neuromodulated Meta-Learning Algorithm (ANML).

90, TITLE:      Safe reinforcement learning for probabilistic reachability and safety specifications: A Lyapunov-based approach
http://arxiv.org/abs/2002.10126
AUTHORS:        Subin Huh ; Insoon Yang
HIGHLIGHT:      In this regard, we propose a model-free safety specification method that learns the maximal probability of safe operation by carefully combining probabilistic reachability analysis and safe reinforcement learning (RL).

91, TITLE:      Real-time Kinematic Ground Truth for the Oxford RobotCar Dataset
http://arxiv.org/abs/2002.10152
AUTHORS:        Will Maddern ; Geoffrey Pascoe ; Matthew Gadd ; Dan Barnes ; Brian Yeomans ; Paul Newman
COMMENTS:       Dataset website: https://robotcar-dataset.robots.ox.ac.uk/
HIGHLIGHT:      We describe the release of reference data towards a challenging long-term localisation and mapping benchmark based on the large-scale Oxford RobotCar Dataset.

92, TITLE:      Exponential Amortized Resource Analysis
http://arxiv.org/abs/2002.09519
AUTHORS:        David M Kahn ; Jan Hoffmann
HIGHLIGHT:      This paper presents and extension of AARA to exponential bounds that preserves the benefits of the technique, such as compositionality and efficient type inference based on linear constraint solving.

93, TITLE:      Fast in-place algorithms for polynomial operations: division, evaluation, interpolation
http://arxiv.org/abs/2002.10304
AUTHORS:        Pascal Giorgi ; Bruno Grenet ; Daniel S. Roche
HIGHLIGHT:      We demonstrate new in-place algorithms for the aforementioned polynomial computations which require only constant extra space and achieve the same asymptotic running time as their out-of-place counterparts.

94, TITLE:      Neuron Shapley: Discovering the Responsible Neurons
http://arxiv.org/abs/2002.09815
AUTHORS:        Amirata Ghorbani ; James Zou
HIGHLIGHT:      We develop Neuron Shapley as a new framework to quantify the contribution of individual neurons to the prediction and performance of a deep network.

95, TITLE:      Learning to Select Bi-Aspect Information for Document-Scale Text Content Manipulation
http://arxiv.org/abs/2002.10210
AUTHORS:        Xiaocheng Feng ; Yawei Sun ; Bing Qin ; Heng Gong ; Yibo Sun ; Wei Bi ; Xiaojiang Liu ; Ting Liu
COMMENTS:       accepted by AAAI2020
HIGHLIGHT:      In this paper, we focus on a new practical task, document-scale text content manipulation, which is the opposite of text style transfer and aims to preserve text styles while altering the content. To tackle those problems, we first build a dataset based on a basketball game report corpus as our testbed, and present an unsupervised neural model with interactive attention mechanism, which is used for learning the semantic relationship between records and reference texts to achieve better content transfer and better style preservation.

96, TITLE:      A Hybrid Approach to Dependency Parsing: Combining Rules and Morphology with Deep Learning
http://arxiv.org/abs/2002.10116
AUTHORS:        Şaziye Betül Özateş ; Arzucan Özgür ; Tunga Güngör ; Balkız Öztürk
COMMENTS:       25 pages, 7 figures
HIGHLIGHT:      We propose two approaches to dependency parsing especially for languages with restricted amount of training data.

97, TITLE:      Predicting Subjective Features from Questions on QA Websites using BERT
http://arxiv.org/abs/2002.10107
AUTHORS:        Issa Annamoradnejad ; Mohammadamin Fazli ; Jafar Habibi
COMMENTS:       5 pages, 4 figures, 2 tables
HIGHLIGHT:      Therefore, with the overall goal of providing solutions for automating moderation actions in Q&A websites, we aim to provide a model to predict 20 quality or subjective aspects of questions in QA websites.

98, TITLE:      GRET: Global Representation Enhanced Transformer
http://arxiv.org/abs/2002.10101
AUTHORS:        Rongxiang Weng ; Haoran Wei ; Shujian Huang ; Heng Yu ; Lidong Bing ; Weihua Luo ; Jiajun Chen
COMMENTS:       Accepted by AAAI 2020
HIGHLIGHT:      In this paper, we propose a novel global representation enhanced Transformer (GRET) to explicitly model global representation in the Transformer network.

99, TITLE:      Low-Resource Knowledge-Grounded Dialogue Generation
http://arxiv.org/abs/2002.10348
AUTHORS:        Xueliang Zhao ; Wei Wu ; Chongyang Tao ; Can Xu ; Dongyan Zhao ; Rui Yan
COMMENTS:       Published in ICLR 2020
HIGHLIGHT:      Yet knowledge-grounded dialogues, as training data for learning such a response generation model, are difficult to obtain.

100, TITLE:     Semi-Supervised Speech Recognition via Local Prior Matching
http://arxiv.org/abs/2002.10336
AUTHORS:        Wei-Ning Hsu ; Ann Lee ; Gabriel Synnaeve ; Awni Hannun
HIGHLIGHT:      In this work, we propose local prior matching (LPM), a semi-supervised objective that distills knowledge from a strong prior (e.g. a language model) to provide learning signal to a discriminative model trained on unlabeled speech.

101, TITLE:     Fixed Encoder Self-Attention Patterns in Transformer-Based Machine Translation
http://arxiv.org/abs/2002.10260
AUTHORS:        Alessandro Raganato ; Yves Scherrer ; Jörg Tiedemann
HIGHLIGHT:      In this paper, we propose to replace all but one attention head of each encoder layer with fixed -- non-learnable -- attentive patterns that are solely based on position and do not require any external knowledge.

102, TITLE:     Multilingual Twitter Corpus and Baselines for Evaluating Demographic Bias in Hate Speech Recognition
http://arxiv.org/abs/2002.10361
AUTHORS:        Xiaolei Huang ; Linzi Xing ; Franck Dernoncourt ; Michael J. Paul
COMMENTS:       Accepted at LREC 2020
HIGHLIGHT:      In this work, we assemble and publish a multilingual Twitter corpus for the task of hate speech detection with inferred four author demographic factors: age, country, gender and race/ethnicity.

103, TITLE:     Improving BERT Fine-Tuning via Self-Ensemble and Self-Distillation
http://arxiv.org/abs/2002.10345
AUTHORS:        Yige Xu ; Xipeng Qiu ; Ligao Zhou ; Xuanjing Huang
COMMENTS:       7 pages, 6 figures
HIGHLIGHT:      In this paper, we improve the fine-tuning of BERT with two effective mechanisms: self-ensemble and self-distillation.

104, TITLE:     Maximum Entropy on the Mean: A Paradigm Shift for Regularization in Image Deblurring
http://arxiv.org/abs/2002.10434
AUTHORS:        Gabriel Rioux ; Rustum Choksi ; Tim Hoheisel ; Christopher Scarvelis
COMMENTS:       15 pages, 7 figures
HIGHLIGHT:      We propose an alternative approach, shifting the paradigm towards regularization at the level of the probability distribution on the space of images.

105, TITLE:     A characterization of proportionally representative committees
http://arxiv.org/abs/2002.09598
AUTHORS:        Haris Aziz ; Barton E. Lee
HIGHLIGHT:      We characterize committees satisfying PSC as possible outcomes of the Minimal Demand rule, which generalizes an approach pioneered by Michael Dummett.

106, TITLE:     Shallow2Deep: Indoor Scene Modeling by Single Image Understanding
http://arxiv.org/abs/2002.09790
AUTHORS:        Yinyu Nie ; Shihui Guo ; Jian Chang ; Xiaoguang Han ; Jiahui Huang ; Shi-Min Hu ; Jian Jun Zhang
COMMENTS:       Accepted by Pattern Recognition
HIGHLIGHT:      We present an automatic indoor scene modeling approach using deep features from neural networks.

107, TITLE:     VisionGuard: Runtime Detection of Adversarial Inputs to Perception Systems
http://arxiv.org/abs/2002.09792
AUTHORS:        Yiannis Kantaros ; Taylor Carpenter ; Sangdon Park ; Radoslav Ivanov ; Sooyong Jang ; Insup Lee ; James Weimer
HIGHLIGHT:      In this paper, we propose VisionGuard, a novel attack- and dataset-agnostic and computationally-light defense mechanism for adversarial inputs to DNN-based perception systems.

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

109, TITLE:     The Knowledge Graph Track at OAEI -- Gold Standards, Baselines, and the Golden Hammer Bias
http://arxiv.org/abs/2002.10283
AUTHORS:        Sven Hertling ; Heiko Paulheim
HIGHLIGHT:      In this paper, we discuss the design of the track and two different strategies of gold standard creation.

110, TITLE:     Automata for Hyperlanguages
http://arxiv.org/abs/2002.09877
AUTHORS:        Borzoo Bonakdarpour ; Sarai Sheinvald
COMMENTS:       12 pages of main paper and another 10 pages appendix
HIGHLIGHT:      We introduce hyperautomata for em hyperlanguages, which are languages over sets of words.

111, TITLE:     A Critical View of the Structural Causal Model
http://arxiv.org/abs/2002.10007
AUTHORS:        Tomer Galanti ; Ofir Nabati ; Lior Wolf
HIGHLIGHT:      In the multivariate case, where one can ensure that the complexities of the cause and effect are balanced, we propose a new adversarial training method that mimics the disentangled structure of the causal model.

112, TITLE:     NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Aggregated Convolutional Feature Maps
http://arxiv.org/abs/2002.10003
AUTHORS:        Maximilian Seitzer
COMMENTS:       Disentanglement Challenge - 33rd Conference on Neural Information Processing Systems (NeurIPS) - NeurIPS 2019
HIGHLIGHT:      This report to our stage 1 submission to the NeurIPS 2019 disentanglement challenge presents a simple image preprocessing method for training VAEs leading to improved disentanglement compared to directly using the images.

113, TITLE:     Optimal strategies in the Fighting Fantasy gaming system: influencing stochastic dynamics by gambling with limited resource
http://arxiv.org/abs/2002.10172
AUTHORS:        Iain G. Johnston
COMMENTS:       Keyword: stochastic game; Markov decision problem; stochastic simulation; dynamic programming; resource allocation; stochastic optimal control; Bellman equation
HIGHLIGHT:      Here, we combine stochastic analysis and simulation with dynamic programming to characterise the dynamical behaviour of the system in the absence and presence of gambling policy.

114, TITLE:     Verifying Array Manipulating Programs with Full-Program Induction
http://arxiv.org/abs/2002.09857
AUTHORS:        Supratik Chakraborty ; Ashutosh Gupta ; Divyesh Unadkat
HIGHLIGHT:      We present a full-program induction technique for proving (a sub-class of) quantified as well as quantifier-free properties of programs manipulating arrays of parametric size N. Instead of inducting over individual loops, our technique inducts over the entire program (possibly containing multiple loops) directly via the program parameter N. Significantly, this does not require generation or use of loop-specific invariants.

115, TITLE:     Notes on neighborhood semantics for logics of unknown truths and false beliefs
http://arxiv.org/abs/2002.09622
AUTHORS:        Jie Fan
COMMENTS:       21 pages
HIGHLIGHT:      In this article, we study logics of unknown truths and false beliefs under neighborhood semantics.

116, TITLE:     Resources for Turkish Dependency Parsing: Introducing the BOUN Treebank and the BoAT Annotation Tool
http://arxiv.org/abs/2002.10416
AUTHORS:        Utku Türk ; Furkan Atmaca ; Şaziye Betül Özateş ; Gözde Berk ; Seyyit Talha Bedir ; Abdullatif Köksal ; Balkız Öztürk Başaran ; Tunga Güngör ; Arzucan Özgür
COMMENTS:       29 pages, 5 figures, 10 tables, submitted to Language Resources and Evaluation
HIGHLIGHT:      In this paper, we describe our contributions and efforts to develop Turkish resources, which include a new treebank (BOUN Treebank) with novel sentences, along with the guidelines we adopted and a new annotation tool we developed (BoAT).

117, TITLE:     The Pragmatic Turn in Explainable Artificial Intelligence (XAI)
http://arxiv.org/abs/2002.09595
AUTHORS:        Andrés Páez
HIGHLIGHT:      In this paper I argue that the search for explainable models and interpretable decisions in AI must be reformulated in terms of the broader project of offering a pragmatic and naturalistic account of understanding in AI.

118, TITLE:     Automatic Cost Function Learning with Interpretable Compositional Networks
http://arxiv.org/abs/2002.09811
AUTHORS:        Florian Richoux ; Jean-François Baffier
HIGHLIGHT:      Here we propose a method to automatically learn a cost function of a constraint, given a function deciding if assignments are valid or not.

119, TITLE:     Discriminative Adversarial Search for Abstractive Summarization
http://arxiv.org/abs/2002.10375
AUTHORS:        Thomas Scialom ; Paul-Alexis Dray ; Sylvain Lamprier ; Benjamin Piwowarski ; Jacopo Staiano
HIGHLIGHT:      We introduce a novel approach for sequence decoding, Discriminative Adversarial Search (DAS), which has the desirable properties of alleviating the effects of exposure bias without requiring external metrics.

120, TITLE:     Conceptual Game Expansion
http://arxiv.org/abs/2002.09636
AUTHORS:        Matthew Guzdial ; Mark Riedl
COMMENTS:       14 pages, 5 figures
HIGHLIGHT:      In this paper we instead learn representations of existing games and use these to approximate a search space of novel games.

121, TITLE:     Anatomy-aware 3D Human Pose Estimation in Videos
http://arxiv.org/abs/2002.10322
AUTHORS:        Tianlang Chen ; Chen Fang ; Xiaohui Shen ; Yiheng Zhu ; Zhili Chen ; Jiebo Luo
HIGHLIGHT:      In this work, we propose a new solution for 3D human pose estimation in videos.

122, TITLE:     Guessing State Tracking for Visual Dialogue
http://arxiv.org/abs/2002.10340
AUTHORS:        Wei Pang ; Xiaojie Wang
COMMENTS:       9 pages, 5 figures, Nov. 2019, https://github.com/xubuvd/guesswhat/blob/master/guesswhat_performance_2019.png
HIGHLIGHT:      This paper proposes the guessing state for the guesser, and regards guess as a process with change of guessing state through a dialogue.

123, TITLE:     Suppressing Uncertainties for Large-Scale Facial Expression Recognition
http://arxiv.org/abs/2002.10392
AUTHORS:        Kai Wang ; Xiaojiang Peng ; Jianfei Yang ; Shijian Lu ; Yu Qiao
COMMENTS:       This manuscript has been accepted by CVPR2020
HIGHLIGHT:      To address this problem, this paper proposes a simple yet efficient Self-Cure Network (SCN) which suppresses the uncertainties efficiently and prevents deep networks from over-fitting uncertain facial images.

124, TITLE:     Sketch Less for More: On-the-Fly Fine-Grained Sketch Based Image Retrieval
http://arxiv.org/abs/2002.10310
AUTHORS:        Ayan Kumar Bhunia ; Yongxin Yang ; Timothy M. Hospedales ; Tao Xiang ; Yi-Zhe Song
COMMENTS:       IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2020
HIGHLIGHT:      In this paper, we reformulate the conventional FG-SBIR framework to tackle these challenges, with the ultimate goal of retrieving the target photo with the least number of strokes possible.

125, TITLE:     Comparing View-Based and Map-Based Semantic Labelling in Real-Time SLAM
http://arxiv.org/abs/2002.10342
AUTHORS:        Zoe Landgraf ; Fabian Falck ; Michael Bloesch ; Stefan Leutenegger ; Andrew Davison
COMMENTS:       ICRA 2020
HIGHLIGHT:      Here, we present an experimental framework and comparison which uses real-time height map fusion as an accessible platform for a fair comparison, opening up the route to further systematic research in this area.

126, TITLE:     Sketchformer: Transformer-based Representation for Sketched Structure
http://arxiv.org/abs/2002.10381
AUTHORS:        Leo Sampaio Ferraz Ribeiro ; Tu Bui ; John Collomosse ; Moacir Ponti
COMMENTS:       Accepted for publication at CVPR 2020
HIGHLIGHT:      We report several variants exploring continuous and tokenized input representations, and contrast their performance.

127, TITLE:     Joint Learning of Assignment and Representation for Biometric Group Membership
http://arxiv.org/abs/2002.10363
AUTHORS:        Marzieh Gheisari ; Teddy Furon ; Laurent Amsaleg
HIGHLIGHT:      This paper proposes a framework for group membership protocols preventing the curious but honest server from reconstructing the enrolled biometric signatures and inferring the identity of querying clients.

128, TITLE:     UnMask: Adversarial Detection and Defense Through Robust Feature Alignment
http://arxiv.org/abs/2002.09576
AUTHORS:        Scott Freitas ; Shang-Tse Chen ; Zijie Wang ; Duen Horng Chau
HIGHLIGHT:      To combat these adversarial attacks, we developed UnMask, an adversarial detection and defense framework based on robust feature alignment.

129, TITLE:     Active Lighting Recurrence by Parallel Lighting Analogy for Fine-Grained Change Detection
http://arxiv.org/abs/2002.09663
AUTHORS:        Qian Zhang ; Wei Feng ; Liang Wan ; Fei-Peng Tian ; Xiaowei Wang ; Ping Tan
HIGHLIGHT:      To this end, we propose to use the simple parallel lighting as an analogy model and based on Lambertian law to compose an instant navigation ball for this purpose.

130, TITLE:     Robust Multimodal Brain Tumor Segmentation via Feature Disentanglement and Gated Fusion
http://arxiv.org/abs/2002.09708
AUTHORS:        Cheng Chen ; Qi Dou ; Yueming Jin ; Hao Chen ; Jing Qin ; Pheng-Ann Heng
COMMENTS:       MICCAI 2019
HIGHLIGHT:      We tackle this challenge and propose a novel multimodal segmentation framework which is robust to the absence of imaging modalities.

131, TITLE:     Particle Filter Based Monocular Human Tracking with a 3D Cardbox Model and a Novel Deterministic Resampling Strategy
http://arxiv.org/abs/2002.09554
AUTHORS:        Ziyuan Liu ; Dongheui Lee ; Wolfgang Sepp
HIGHLIGHT:      In this paper, a motion capturing algorithm is proposed for upper body motion tracking.

132, TITLE:     Temporal Sparse Adversarial Attack on Gait Recognition
http://arxiv.org/abs/2002.09674
AUTHORS:        Ziwen He ; Wei Wang ; Jing Dong ; Tieniu Tan
HIGHLIGHT:      In this paper, we demonstrate that the state-of-the-art gait recognition model is vulnerable to adversarial attacks.

133, TITLE:     Communication Contention Aware Scheduling of Multiple Deep Learning Training Jobs
http://arxiv.org/abs/2002.10105
AUTHORS:        Qiang Wang ; Shaohuai Shi ; Canhui Wang ; Xiaowen Chu
HIGHLIGHT:      We thus propose a provable algorithm, AdaDUAL, to efficiently schedule those communication tasks.

134, TITLE:     Learning From Strategic Agents: Accuracy, Improvement, and Causality
http://arxiv.org/abs/2002.10066
AUTHORS:        Yonadav Shavit ; Benjamin Edelman ; Brian Axelrod
COMMENTS:       18 pages
HIGHLIGHT:      As our main contribution, we provide the first algorithms for learning accuracy-optimizing, improvement-optimizing, and causal-precision-optimizing linear regression models directly from data, without prior knowledge of agents' possible actions.

135, TITLE:     Estimating Q(s,s') with Deep Deterministic Dynamics Gradients
http://arxiv.org/abs/2002.09505
AUTHORS:        Ashley D. Edwards ; Himanshu Sahni ; Rosanne Liu ; Jane Hung ; Ankit Jain ; Rui Wang ; Adrien Ecoffet ; Thomas Miconi ; Charles Isbell ; Jason Yosinski
HIGHLIGHT:      In this paper, we introduce a novel form of value function, $Q(s, s')$, that expresses the utility of transitioning from a state $s$ to a neighboring state $s'$ and then acting optimally thereafter.

136, TITLE:     Crossing the Reality Gap with Evolved Plastic Neurocontrollers
http://arxiv.org/abs/2002.09854
AUTHORS:        Huanneng Qiu ; Matthew Garratt ; David Howard ; Sreenatha Anavatti
COMMENTS:       Submitted to GECCO2020
HIGHLIGHT:      Here we try to overcome the transfer problem from a different perspective, by designing a spiking neurocontroller which uses synaptic plasticity to cross the reality gap via online adaptation.

137, TITLE:     Monocular Direct Sparse Localization in a Prior 3D Surfel Map
http://arxiv.org/abs/2002.09923
AUTHORS:        Haoyang Ye ; Huaiyang Huang ; Ming Liu
COMMENTS:       7 pages, 6 figures, to appear in ICRA 2020
HIGHLIGHT:      In this paper, we introduce an approach to tracking the pose of a monocular camera in a prior surfel map.

138, TITLE:     Self-Supervised Poisson-Gaussian Denoising
http://arxiv.org/abs/2002.09558
AUTHORS:        Wesley Khademi ; Sonia Rao ; Clare Minnerath ; Guy Hagen ; Jonathan Ventura
HIGHLIGHT:      We introduce a new training strategy to handle Poisson-Gaussian noise which is the standard noise model for microscope images.



==========Updates to Previous Papers==========
1, TITLE:       Causal query in observational data with hidden variables
http://arxiv.org/abs/2001.10269
AUTHORS:        Debo Cheng ; Jiuyong Li ; Lin Liu ; Jixue Liu ; Kui Yu ; Thuc Duy Le
HIGHLIGHT:      In this paper, we develop theorems for using local search to find a superset of the adjustment (or confounding) variables for causal effect estimation from observational data under a realistic pretreatment assumption.

2, TITLE:       Anchoring Theory in Sequential Stackelberg Games
http://arxiv.org/abs/1912.03564
AUTHORS:        Jan Karwowski ; Jacek Mańdziuk ; Adam Żychowski
HIGHLIGHT:      This paper proposes an efficient formulation of AT in sequential extensive-form SGs (named ATSG), suitable for Mixed-Integer Linear Program (MILP) solution methods.

3, TITLE:       Santha-Vazirani sources, deterministic condensers and very strong extractors
http://arxiv.org/abs/1907.04640
AUTHORS:        Dmitry Gavinsky ; Pavel Pudlák
HIGHLIGHT:      It is intuitively obvious that strong SV-sources are more than just high-min-entropy sources, and this work explores the intuition.

4, TITLE:       AtomNAS: Fine-Grained End-to-End Neural Architecture Search
http://arxiv.org/abs/1912.09640
AUTHORS:        Jieru Mei ; Yingwei Li ; Xiaochen Lian ; Xiaojie Jin ; Linjie Yang ; Alan Yuille ; Jianchao Yang
COMMENTS:       ICLR 2020 camera ready version
HIGHLIGHT:      Based on this search space, we propose a resource-aware architecture search framework which automatically assigns the computational resources (e.g., output channel numbers) for each operation by jointly considering the performance and the computational cost.

5, TITLE:       Unpaired Point Cloud Completion on Real Scans using Adversarial Training
http://arxiv.org/abs/1904.00069
AUTHORS:        Xuelin Chen ; Baoquan Chen ; Niloy J. Mitra
COMMENTS:       ICLR 2020
HIGHLIGHT:      We develop a first approach that works directly on input point clouds, does not require paired training data, and hence can directly be applied to real scans for scan completion.

6, TITLE:       AdvectiveNet: An Eulerian-Lagrangian Fluidic reservoir for Point Cloud Processing
http://arxiv.org/abs/2002.00118
AUTHORS:        Xingzhe He ; Helen Lu Cao ; Bo Zhu
COMMENTS:       ICLR 2020
HIGHLIGHT:      This paper presents a novel physics-inspired deep learning approach for point cloud processing motivated by the natural flow phenomena in fluid mechanics.

7, TITLE:       Blockchain Intelligence: When Blockchain Meets Artificial Intelligence
http://arxiv.org/abs/1912.06485
AUTHORS:        Zibin Zheng ; Hong-Ning Dai
COMMENTS:       5 pages, 5 figures
HIGHLIGHT:      This article presents an introduction of the convergence of blockchain and artificial intelligence (namely blockchain intelligence).

8, TITLE:       PDA: Progressive Data Augmentation for General Robustness of Deep Neural Networks
http://arxiv.org/abs/1909.04839
AUTHORS:        Hang Yu ; Aishan Liu ; Xianglong Liu ; Gengchao Li ; Ping Luo ; Ran Cheng ; Jichen Yang ; Chongzhi Zhang
HIGHLIGHT:      To address this problem, we propose a simple yet effective method, named Progressive Data Augmentation (PDA), which enables general robustness of DNNs by progressively injecting diverse adversarial noises during training.

9, TITLE:       Learning Perception and Planning with Deep Active Inference
http://arxiv.org/abs/2001.11841
AUTHORS:        Ozan Çatal ; Tim Verbelen ; Johannes Nauta ; Cedric De Boom ; Bart Dhoedt
COMMENTS:       Accepted on ICASSP 2020
HIGHLIGHT:      In this paper we use recent advances in deep learning to learn the state space and approximate the necessary probability distributions to engage in active inference.

10, TITLE:      Distributional Soft Actor-Critic: Off-Policy Reinforcement Learning for Addressing Value Estimation Errors
http://arxiv.org/abs/2001.02811
AUTHORS:        Jingliang Duan ; Yang Guan ; Shengbo Eben Li ; Yangang Ren ; Bo Cheng
HIGHLIGHT:      We evaluate our method on the suite of MuJoCo continuous control tasks, achieving state-of-the-art performance.

11, TITLE:      On Defending Against Label Flipping Attacks on Malware Detection Systems
http://arxiv.org/abs/1908.04473
AUTHORS:        Rahim Taheri ; Reza Javidan ; Mohammad Shojafar ; Zahra Pooranian ; Ali Miri ; Mauro Conti
COMMENTS:       19 pages, 6 figures, 2 tables, NCAA Springer Journal
HIGHLIGHT:      In this paper, we design an architecture to tackle the Android malware detection problem in IoT systems.

12, TITLE:      Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers
http://arxiv.org/abs/1912.03277
AUTHORS:        Divyat Mahajan ; Chenhao Tan ; Amit Sharma
COMMENTS:       2019 NeurIPS Workshop on Do the right thing: Machine learning and Causal Inference for improved decision making
HIGHLIGHT:      When feasibility constraints may not be easily expressed, we propose an alternative method that optimizes for feasibility as people interact with its output and provide oracle-like feedback.

13, TITLE:      Trace-Relating Compiler Correctness and Secure Compilation
http://arxiv.org/abs/1907.05320
AUTHORS:        Carmine Abate ; Roberto Blanco ; Stefan Ciobaca ; Adrien Durier ; Deepak Garg ; Catalin Hritcu ; Marco Patrignani ; Éric Tanter ; Jérémy Thibault
COMMENTS:       ESOP'20 camera ready version together with online appendix
HIGHLIGHT:      To overcome this issue, we study a generalized compiler correctness definition, which uses source and target traces drawn from potentially different sets and connected by an arbitrary relation.

14, TITLE:      A Comparison of the Taguchi Method and Evolutionary Optimization in Multivariate Testing
http://arxiv.org/abs/1808.08347
AUTHORS:        Jingbo Jiang ; Diego Legrand ; Robert Severn ; Risto Miikkulainen
COMMENTS:       5 pages, 4 figures, IAAI-19
HIGHLIGHT:      In contrast to the standard A/B testing, multivariate approach aims at evaluating a large number of values in a few key variables systematically.

15, TITLE:      Correctness of Automatic Differentiation via Diffeologies and Categorical Gluing
http://arxiv.org/abs/2001.02209
AUTHORS:        Mathieu Huot ; Sam Staton ; Matthijs Vákár
COMMENTS:       Proceedings of FoSSaCS 2020
HIGHLIGHT:      We present semantic correctness proofs of Automatic Differentiation (AD).

16, TITLE:      Creative AI Through Evolutionary Computation
http://arxiv.org/abs/1901.03775
AUTHORS:        Risto Miikkulainen
HIGHLIGHT:      The main power of artificial intelligence is not in modeling what we already know, but in creating solutions that are new.

17, TITLE:      Naive Gabor Networks for Hyperspectral Image Classification
http://arxiv.org/abs/1912.03991
AUTHORS:        Chenying Liu ; Jun Li ; Lin He ; Antonio J. Plaza ; Shutao Li ; Bo Li
COMMENTS:       This paper has been accepted by IEEE TNNLS
HIGHLIGHT:      To address these problems, in this paper, we introduce naive Gabor Networks or Gabor-Nets which, for the first time in the literature, design and learn CNN kernels strictly in the form of Gabor filters, aiming to reduce the number of involved parameters and constrain the solution space, and hence improve the performances of CNNs.

18, TITLE:      Single Image Super-resolution via Dense Blended Attention Generative Adversarial Network for Clinical Diagnosis
http://arxiv.org/abs/1906.06575
AUTHORS:        Kewen Liu ; Yuan Ma ; Hongxia Xiong ; Zejun Yan ; Zhijun Zhou ; Chaoyang Liu ; Panpan Fang ; Xiaojun Li ; Yalei Chen
COMMENTS:       We abandoned this paper due to its limitation only applied on medical images, please view our lastest work at arXiv:1911.03464
HIGHLIGHT:      Single Image Super-resolution via Dense Blended Attention Generative Adversarial Network for Clinical Diagnosis

19, TITLE:      Goal-constrained Planning Domain Model Verification of Safety Properties
http://arxiv.org/abs/1811.09231
AUTHORS:        Anas Shrinah ; Kerstin Eder
HIGHLIGHT:      In this paper, we discuss the downside of unconstrained planning domain model verification.

20, TITLE:      ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training
http://arxiv.org/abs/2001.04063
AUTHORS:        Yu Yan ; Weizhen Qi ; Yeyun Gong ; Dayiheng Liu ; Nan Duan ; Jiusheng Chen ; Ruofei Zhang ; Ming Zhou
HIGHLIGHT:      In this paper, we present a new sequence-to-sequence pre-training model called ProphetNet, which introduces a novel self-supervised objective named future n-gram prediction and the proposed n-stream self-attention mechanism.

21, TITLE:      Long-Term Human Video Generation of Multiple Futures Using Poses
http://arxiv.org/abs/1904.07538
AUTHORS:        Naoya Fushishita ; Antonio Tejero-de-Pablos ; Yusuke Mukuta ; Tatsuya Harada
HIGHLIGHT:      In this paper, we propose a novel method for future human pose prediction capable of predicting multiple long-term futures.

22, TITLE:      An Explicit Local and Global Representation Disentanglement Framework with Applications in Deep Clustering and Unsupervised Object Detection
http://arxiv.org/abs/2001.08957
AUTHORS:        Rujikorn Charakorn ; Yuttapong Thawornwattana ; Sirawaj Itthipuripat ; Nick Pawlowski ; Poramate Manoonpong ; Nat Dilokthanakul
COMMENTS:       13 pages and 9 figures
HIGHLIGHT:      In this work, we propose a framework, called SPLIT, which allows us to disentangle local and global information into two separate sets of latent variables within the variational autoencoder (VAE) framework.

23, TITLE:      GFF: Gated Fully Fusion for Semantic Segmentation
http://arxiv.org/abs/1904.01803
AUTHORS:        Xiangtai Li ; Houlong Zhao ; Lei Han ; Yunhai Tong ; Kuiyuan Yang
COMMENTS:       accepted by AAAI-2020(oral)
HIGHLIGHT:      In this paper, we propose a new architecture, named Gated Fully Fusion (GFF), to selectively fuse features from multiple levels using gates in a fully connected way.

24, TITLE:      Improved Res2Net model for Person re-identification
http://arxiv.org/abs/1910.04061
AUTHORS:        Zongjing Cao ; Hyo Jong Lee
COMMENTS:       6 pages
HIGHLIGHT:      In this paper, we propose a multi-task network base on an improved Res2Net model that simultaneously computes the identification loss and verification loss of two pedestrian images.

25, TITLE:      Deep Multi-Facial patches Aggregation Network for Expression Classification from Face Images
http://arxiv.org/abs/1909.10305
AUTHORS:        Amine Djerghri ; Ahmed Rachid Hazourli ; Alice Othmani
COMMENTS:       we have a new version of the paper arXiv:2002.09298
HIGHLIGHT:      In this paper, we investigate HCI via a deep multi-facial patches aggregation network for Face Expression Recognition (FER).

26, TITLE:      Human-Centered Artificial Intelligence: Reliable, Safe & Trustworthy
http://arxiv.org/abs/2002.04087
AUTHORS:        Ben Shneiderman
COMMENTS:       15 pages, followed by 4 pages of references, 6 figures
HIGHLIGHT:      The Human-Centered Artificial Intelligence (HCAI) framework clarifies how to (1) design for high levels of human control and high levels of computer automation so as to increase human performance, (2) understand the situations in which full human control or full computer control are necessary, and (3) avoid the dangers of excessive human control or excessive computer control.

27, TITLE:      iCap: Interactive Image Captioning with Predictive Text
http://arxiv.org/abs/2001.11782
AUTHORS:        Zhengxiong Jia ; Xirong Li
HIGHLIGHT:      In this paper we study a brand new topic of interactive image captioning with human in the loop.

28, TITLE:      Coresets for Data-efficient Training of Machine Learning Models
http://arxiv.org/abs/1906.01827
AUTHORS:        Baharan Mirzasoleiman ; Jeff Bilmes ; Jure Leskovec
HIGHLIGHT:      Here we develop CRAIG, a method to select a weighted subset (or coreset) of training data that closely estimates the full gradient by maximizing a submodular function.

29, TITLE:      Provable Self-Play Algorithms for Competitive Reinforcement Learning
http://arxiv.org/abs/2002.04017
AUTHORS:        Yu Bai ; Chi Jin
HIGHLIGHT:      We introduce a self-play algorithm---Value Iteration with Upper/Lower Confidence Bound (VI-ULCB), and show that it achieves regret $\mathcal{\tilde{O}}(\sqrt{T})$ after playing $T$ steps of the game.

30, TITLE:      No-regret Exploration in Contextual Reinforcement Learning
http://arxiv.org/abs/1903.06187
AUTHORS:        Aditya Modi ; Ambuj Tewari
COMMENTS:       Under review. 25 pages, 2 figures
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 models (GLMs).

31, TITLE:      Dueling Posterior Sampling for Preference-Based Reinforcement Learning
http://arxiv.org/abs/1908.01289
AUTHORS:        Ellen R. Novoseller ; Yibing Wei ; Yanan Sui ; Yisong Yue ; Joel W. Burdick
COMMENTS:       9 pages before references and Appendix; 46 pages total; 7 figures; 4 tables. This replacement corrects a theoretical limitation in the previous version, adds additional simulation results, and extends the empirical analysis
HIGHLIGHT:      Building upon ideas from preference-based bandit learning and posterior sampling in RL, we present DUELING POSTERIOR SAMPLING (DPS), which employs preference-based posterior sampling to learn both the system dynamics and the underlying utility function that governs the preference feedback.

32, TITLE:      BLK-REW: A Unified Block-based DNN Pruning Framework using Reweighted Regularization Method
http://arxiv.org/abs/2001.08357
AUTHORS:        Xiaolong Ma ; Zhengang Li ; Yifan Gong ; Tianyun Zhang ; Wei Niu ; Zheng Zhan ; Pu Zhao ; Jian Tang ; Xue Lin ; Bin Ren ; Yanzhi Wang
HIGHLIGHT:      To solve the problem, we propose a new block-based pruning framework that comprises a general and flexible structured pruning dimension as well as a powerful and efficient reweighted regularization method.

33, TITLE:      TarMAC: Targeted Multi-Agent Communication
http://arxiv.org/abs/1810.11187
AUTHORS:        Abhishek Das ; Théophile Gervet ; Joshua Romoff ; Dhruv Batra ; Devi Parikh ; Michael Rabbat ; Joelle Pineau
COMMENTS:       ICML 2019
HIGHLIGHT:      We propose a targeted communication architecture for multi-agent reinforcement learning, where agents learn both what messages to send and whom to address them to while performing cooperative tasks in partially-observable environments.

34, TITLE:      Differentiable Product Quantization for End-to-End Embedding Compression
http://arxiv.org/abs/1908.09756
AUTHORS:        Ting Chen ; Lala Li ; Yizhou Sun
HIGHLIGHT:      In this work, we propose a generic and end-to-end learnable compression framework termed differentiable product quantization (DPQ).

35, TITLE:      Semi-Supervised Class Discovery
http://arxiv.org/abs/2002.03480
AUTHORS:        Jeremy Nixon ; Jeremiah Liu ; David Berthelot
HIGHLIGHT:      We introduce the Dataset Reconstruction Accuracy, a new and important measure of the effectiveness of a model's ability to create labels. We introduce benchmarks against this Dataset Reconstruction metric.

36, TITLE:      Learning Fitness Functions for Genetic Algorithms
http://arxiv.org/abs/1908.08783
AUTHORS:        Shantanu Mandal ; Todd A. Anderson ; Javier S. Turek ; Justin Gottschlich ; Shengtian Zhou ; Abdullah Muzahid
HIGHLIGHT:      In this work, we propose a framework based on genetic algorithms to solve this problem.

37, TITLE:      Enabling Multi-Shell b-Value Generalizability of Data-Driven Diffusion Models with Deep SHORE
http://arxiv.org/abs/1907.06319
AUTHORS:        Vishwesh Nath ; Ilwoo Lyu ; Kurt G. Schilling ; Prasanna Parvathaneni ; Colin B. Hansen ; Yucheng Tang ; Yuankai Huo ; Vaibhav A. Janve ; Yurui Gao ; Iwona Stepniewska ; Adam W. Anderson ; Bennett A. Landman
HIGHLIGHT:      To ensure consistency of hyper-parameter optimization for SHORE, we present our Deep SHORE approach to learn on a data-optimized manifold.

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

39, TITLE:      Adaptive and Compressive Beamforming Using Deep Learning for Medical Ultrasound
http://arxiv.org/abs/1907.10257
AUTHORS:        Shujaat Khan ; Jaeyoung Huh ; Jong Chul Ye
COMMENTS:       This is a significantly extended version of the original paper in arXiv:1901.01706. This paper is accepted for IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
HIGHLIGHT:      To address this problem, here we propose a deep learning-based beamformer to generate significantly improved images over widely varying measurement conditions and channel subsampling patterns.

40, TITLE:      Compositionality decomposed: how do neural networks generalise?
http://arxiv.org/abs/1908.08351
AUTHORS:        Dieuwke Hupkes ; Verna Dankers ; Mathijs Mul ; Elia Bruni
HIGHLIGHT:      As a response to this controversy, we present a set of tests that provide a bridge between, on the one hand, the vast amount of linguistic and philosophical theory about compositionality of language and, on the other, the successful neural models of language.

41, TITLE:      Relation Adversarial Network for Low Resource Knowledge Graph Completion
http://arxiv.org/abs/1911.03091
AUTHORS:        Ningyu Zhang ; Shumin Deng ; Zhanlin Sun ; Jiaoayan Chen ; Wei Zhang ; Huajun Chen
COMMENTS:       WWW2020
HIGHLIGHT:      In this work, we aim at predicting new facts under a challenging setting where only limited training instances are available.

42, TITLE:      Establishing Strong Baselines for the New Decade: Sequence Tagging, Syntactic and Semantic Parsing with BERT
http://arxiv.org/abs/1908.04943
AUTHORS:        Han He ; Jinho D. Choi
HIGHLIGHT:      This paper presents new state-of-the-art models for three tasks, part-of-speech tagging, syntactic parsing, and semantic parsing, using the cutting-edge contextualized embedding framework known as BERT.

43, TITLE:      Self-Training for End-to-End Speech Recognition
http://arxiv.org/abs/1909.09116
AUTHORS:        Jacob Kahn ; Ann Lee ; Awni Hannun
COMMENTS:       To be published in the 45th IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2020
HIGHLIGHT:      We revisit self-training in the context of end-to-end speech recognition.

44, TITLE:      How Much Knowledge Can You Pack Into the Parameters of a Language Model?
http://arxiv.org/abs/2002.08910
AUTHORS:        Adam Roberts ; Colin Raffel ; Noam Shazeer
HIGHLIGHT:      In this short paper, we measure the practical utility of this approach by fine-tuning pre-trained models to answer questions without access to any external context or knowledge.

45, TITLE:      Back-Projection based Fidelity Term for Ill-Posed Linear Inverse Problems
http://arxiv.org/abs/1906.06794
AUTHORS:        Tom Tirer ; Raja Giryes
HIGHLIGHT:      In this paper, we examine a different fidelity term, which has been implicitly used by the recently proposed iterative denoising and backward projections (IDBP) framework.

46, TITLE:      Under the Radar: Learning to Predict Robust Keypoints for Odometry Estimation and Metric Localisation in Radar
http://arxiv.org/abs/2001.10789
AUTHORS:        Dan Barnes ; Ingmar Posner
COMMENTS:       Video summary: https://youtu.be/L-PO7nxWpJU
HIGHLIGHT:      This paper presents a self-supervised framework for learning to detect robust keypoints for odometry estimation and metric localisation in radar.

47, TITLE:      Attentive Group Equivariant Convolutional Networks
http://arxiv.org/abs/2002.03830
AUTHORS:        David W. Romero ; Erik J. Bekkers ; Jakub M. Tomczak ; Mark Hoogendoorn
HIGHLIGHT:      In this paper, we present attentive group equivariant convolutions, a generalization of the group convolution, in which attention is applied during the course of convolution to accentuate meaningful symmetry combinations and suppress non-plausible, misleading ones.

48, TITLE:      Hope Speech Detection: A Computational Analysis of the Voice of Peace
http://arxiv.org/abs/1909.12940
AUTHORS:        Shriphani Palakodety ; Ashiqur R. KhudaBukhsh ; Jaime G. Carbonell
COMMENTS:       Minor edits
HIGHLIGHT:      In this paper, we analyze this evolving international crisis via a substantial corpus constructed using comments on YouTube videos (921,235 English comments posted by 392,460 users out of 2.04 million overall comments by 791,289 users on 2,890 videos).

49, TITLE:      Post-training Quantization with Multiple Points: Mixed Precision without Mixed Precision
http://arxiv.org/abs/2002.09049
AUTHORS:        Xingchao Liu ; Mao Ye ; Dengyong Zhou ; Qiang Liu
HIGHLIGHT:      We propose multipoint quantization, a quantization method that approximates a full-precision weight vector using a linear combination of multiple vectors of low-bit numbers; this is in contrast to typical quantization methods that approximate each weight using a single low precision number.

50, TITLE:      Denoising and Regularization via Exploiting the Structural Bias of Convolutional Generators
http://arxiv.org/abs/1910.14634
AUTHORS:        Reinhard Heckel ; Mahdi Soltanolkotabi
COMMENTS:       final ICRL version; simplifications in the proof
HIGHLIGHT:      In this paper, we attribute this effect to a particular architectural choice of convolutional networks, namely convolutions with fixed interpolating filters.

51, TITLE:      Better Long-Range Dependency By Bootstrapping A Mutual Information Regularizer
http://arxiv.org/abs/1905.11978
AUTHORS:        Yanshuai Cao ; Peng Xu
COMMENTS:       Camera-ready for AISTATS 2020
HIGHLIGHT:      In this work, we develop a novel regularizer to improve the learning of long-range dependency of sequence data.

52, TITLE:      Unsupervised Attention Mechanism across Neural Network Layers
http://arxiv.org/abs/1902.10658
AUTHORS:        Baihan Lin
HIGHLIGHT:      Inspired by the adaptation phenomenon of neuronal firing, we propose an unsupervised attention mechanism (UAM) which computes the statistical regularity in the implicit space of neural networks under the Minimum Description Length (MDL) principle.

53, TITLE:      Testing Scenario Library Generation for Connected and Automated Vehicles, Part II: Case Studies
http://arxiv.org/abs/1905.03428
AUTHORS:        Shuo Feng ; Yiheng Feng ; Haowei Sun ; Shao Bao ; Yi Zhang ; Henry X. Liu
COMMENTS:       12 pages, 13 figures
HIGHLIGHT:      Three typical cases, including cut-in, highway-exit, and car-following, are designed and studied in this paper.

54, TITLE:      A method for deriving information from running R code
http://arxiv.org/abs/2002.07472
AUTHORS:        Mark P. J. van der Loo
COMMENTS:       11 pages, 1 figure. Accepted for publication by the R Journal (2020-02-20)
HIGHLIGHT:      In this paper we demonstrate an approach that abstracts collection and processing of such secondary information from the running R script.

55, 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.

56, TITLE:      Node Max-Cut and Computing Equilibria in Linear Weighted Congestion Games
http://arxiv.org/abs/1911.08704
AUTHORS:        Dimitris Fotakis ; Vardis Kandiros ; Thanasis Lianeas ; Nikos Mouzakis ; Panagiotis Patsilinakos ; Stratis Skoulakis
HIGHLIGHT:      In this work, we seek a more refined understanding of the complexity of local optimum computation for Max-Cut and pure Nash equilibrium (PNE) computation for congestion games with weighted players and linear latency functions.

57, TITLE:      FrameAxis: Characterizing Framing Bias and Intensity with Word Embedding
http://arxiv.org/abs/2002.08608
AUTHORS:        Haewoon Kwak ; Jisun An ; Yong-Yeol Ahn
COMMENTS:       8 pages; typos corrected (email address)
HIGHLIGHT:      We propose FrameAxis, a method of characterizing the framing of a given text by identifying the most relevant semantic axes ("microframes") defined by antonym word pairs.

58, TITLE:      A Multilingual View of Unsupervised Machine Translation
http://arxiv.org/abs/2002.02955
AUTHORS:        Xavier Garcia ; Pierre Foret ; Thibault Sellam ; Ankur P. Parikh
COMMENTS:       Added new reference, fixed typos
HIGHLIGHT:      We present a probabilistic framework for multilingual neural machine translation that encompasses supervised and unsupervised setups, focusing on unsupervised translation.

59, TITLE:      Hard and Easy Instances of L-Tromino Tilings
http://arxiv.org/abs/1710.04640
AUTHORS:        Javier T. Akagi ; Carlos F. Gaona ; Fabricio Mendoza ; Manjil P. Saikia ; Marcos Villagra
COMMENTS:       Full extended version of LNCS 11355:82-95 (WALCOM 2019)
HIGHLIGHT:      We study tilings of regions in the square lattice with L-shaped trominoes.

60, TITLE:      3D Semi-Supervised Learning with Uncertainty-Aware Multi-View Co-Training
http://arxiv.org/abs/1811.12506
AUTHORS:        Yingda Xia ; Fengze Liu ; Dong Yang ; Jinzheng Cai ; Lequan Yu ; Zhuotun Zhu ; Daguang Xu ; Alan Yuille ; Holger Roth
COMMENTS:       Accepted to WACV 2020
HIGHLIGHT:      In this paper, we propose a novel framework, uncertainty-aware multi-view co-training (UMCT), to address semi-supervised learning on 3D data, such as volumetric data from medical imaging.

61, TITLE:      An Overview of Perception and Decision-Making in Autonomous Systems in the Era of Learning
http://arxiv.org/abs/2001.02319
AUTHORS:        Yang Tang ; Chaoqiang Zhao ; Jianrui Wang ; Chongzhen Zhang ; Qiyu Sun ; Weixing Zheng ; Wenli Du ; Feng Qian ; Juergen Kurths
COMMENTS:       27 pages, 8 figures
HIGHLIGHT:      In this review, we focus on the applications of learning-based approaches in perception and decision-making in autonomous systems, which is different from previous reviews that discussed traditional methods.

62, TITLE:      PolarMask: Single Shot Instance Segmentation with Polar Representation
http://arxiv.org/abs/1909.13226
AUTHORS:        Enze Xie ; Peize Sun ; Xiaoge Song ; Wenhai Wang ; Ding Liang ; Chunhua Shen ; Ping Luo
COMMENTS:       Accepted to Proc. IEEE Conf. Computer Vision and Pattern Recognition 2020
HIGHLIGHT:      In this paper, we introduce an anchor-box free and single shot instance segmentation method, which is conceptually simple, fully convolutional and can be used as a mask prediction module for instance segmentation, by easily embedding it into most off-the-shelf detection methods.

63, TITLE:      Multi-Task Learning by a Top-Down Control Network
http://arxiv.org/abs/2002.03335
AUTHORS:        Hila Levi ; Shimon Ullman
HIGHLIGHT:      We propose a novel architecture that uses a top-down network to modify the main network according to the task in a channel-wise, as well as spatial-wise, image-dependent computation scheme.

64, TITLE:      Deep Matrix Factorization with Spectral Geometric Regularization
http://arxiv.org/abs/1911.07255
AUTHORS:        Amit Boyarski ; Sanketh Vedula ; Alex Bronstein
HIGHLIGHT:      We propose an explicit spectral regularization scheme that is able to make DMF models competitive on real benchmarks, while still maintaining the implicit regularization induced by gradient descent, thus enjoying the best of both worlds.

65, TITLE:      A Modular Termination Method for Second-Order Computation
http://arxiv.org/abs/1912.03434
AUTHORS:        Makoto Hamana
COMMENTS:       23 pages
HIGHLIGHT:      We present a new modular proof method of termination for second-order computation, and report its implementation SOL.

66, TITLE:      An empirical study of Conv-TasNet
http://arxiv.org/abs/2002.08688
AUTHORS:        Berkan Kadioglu ; Michael Horgan ; Xiaoyu Liu ; Jordi Pons ; Dan Darcy ; Vivek Kumar
COMMENTS:       In proceedings of ICASSP2020
HIGHLIGHT:      In this paper, we conduct an empirical study of Conv-TasNet and propose an enhancement to the encoder/decoder that is based on a (deep) non-linear variant of it.

67, TITLE:      Synchronous Transformers for End-to-End Speech Recognition
http://arxiv.org/abs/1912.02958
AUTHORS:        Zhengkun Tian ; Jiangyan Yi ; Ye Bai ; Jianhua Tao ; Shuai Zhang ; Zhengqi Wen
COMMENTS:       Accepted by ICASSP 2020
HIGHLIGHT:      In this paper, we propose a model named synchronous transformer to address this problem, which can predict the output sequence chunk by chunk.

68, TITLE:      Pano Popups: Indoor 3D Reconstruction with a Plane-Aware Network
http://arxiv.org/abs/1907.00939
AUTHORS:        Marc Eder ; Pierre Moulon ; Li Guan
COMMENTS:       2019 International Conference on 3D Vision (3DV). IEEE, 2019
HIGHLIGHT:      In this work we present a method to train a plane-aware convolutional neural network for dense depth and surface normal estimation as well as plane boundaries from a single indoor $360^\circ$ image.

69, TITLE:      Nanoscale Microscopy Images Colorization Using Neural Networks
http://arxiv.org/abs/1912.07964
AUTHORS:        Israel Goytom ; Qin Wang ; Tianxiang Yu ; Kunjie Dai ; Kris Sankaran ; Xinfei Zhou ; Dongdong Lin
COMMENTS:       27 pages, 6 figures
HIGHLIGHT:      In this paper, inspired by the achievement of machine learning algorithms on different science fields, we introduce two artificial neural networks for gray microscopy image colorization: An end-to-end convolutional neural network (CNN) with a pre-trained model for feature extraction and a pixel-to-pixel neural style transfer convolutional neural network (NST-CNN), which can colorize gray microscopy images with semantic information learned from a user-provided colorful image at inference time.

70, TITLE:      Learning Numerical Observers using Unsupervised Domain Adaptation
http://arxiv.org/abs/2002.03763
AUTHORS:        Shenghua He ; Weimin Zhou ; Hua Li ; Mark A. Anastasio
COMMENTS:       SPIE Medical Imaging 2020 (Oral)
HIGHLIGHT:      In this study, we propose and investigate the use of an adversarial domain adaptation method to mitigate the deleterious effects of domain shift between simulated and experimental image data for deep learning-based numerical observers (DL-NOs) that are trained on simulated images but applied to experimental ones.

71, TITLE:      Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving
http://arxiv.org/abs/1812.07179
AUTHORS:        Yan Wang ; Wei-Lun Chao ; Divyansh Garg ; Bharath Hariharan ; Mark Campbell ; Kilian Q. Weinberger
COMMENTS:       Accepted by CVPR 2019
HIGHLIGHT:      Taking the inner workings of convolutional neural networks into consideration, we propose to convert image-based depth maps to pseudo-LiDAR representations --- essentially mimicking the LiDAR signal.

72, TITLE:      Multimodal Representation Model based on Graph-Based Rank Fusion
http://arxiv.org/abs/1912.10314
AUTHORS:        Icaro Cavalcante Dourado ; Salvatore Tabbone ; Ricardo da Silva Torres
HIGHLIGHT:      This paper proposes an unsupervised representation model, based on rank-fusion graphs, for general applicability in multimodal tasks, either unsupervised or supervised prediction.

73, TITLE:      A convolutional neural network reaches optimal sensitivity for detecting some, but not all, patterns
http://arxiv.org/abs/1911.05055
AUTHORS:        Fabian Reith ; Brian Wandell
COMMENTS:       25 pages, 10 figures, pre-print
HIGHLIGHT:      We investigate the performance of a convolutional neural network (CNN) at detecting a signal-known-exactly in Poisson noise.

74, TITLE:      Auto-ML Deep Learning for Rashi Scripts OCR
http://arxiv.org/abs/1811.01290
AUTHORS:        Shahar Mahpod ; Yosi Keller
COMMENTS:       The paper is under consideration at Pattern Recognition Letters
HIGHLIGHT:      In this work we propose an OCR scheme for manuscripts printed in Rashi font that is an ancient Hebrew font and corresponding dialect used in religious Jewish literature, for more than 600 years.

75, TITLE:      The DIDI dataset: Digital Ink Diagram data
http://arxiv.org/abs/2002.09303
AUTHORS:        Philippe Gervais ; Thomas Deselaers ; Emre Aksan ; Otmar Hilliges
HIGHLIGHT:      The dataset was obtained using a prompted data collection effort. We are releasing a dataset of diagram drawings with dynamic drawing information.

76, TITLE:      On The Radon--Nikodym Spectral Approach With Optimal Clustering
http://arxiv.org/abs/1906.00460
AUTHORS:        Vladislav Gennadievich Malyshkin
COMMENTS:       Relation to PCA variation expansion is added. Whereas a regular PCA variation expansion depends on attributes normalizing, the PCA variation expansion in the Lebesgue quadrature arXiv:1807.06007 basis is unique thus does not depend on attributes scale, moreover it is invariant relatively any non-degenerated linear transform of input vector components. Christoffel function solution to vector label
HIGHLIGHT:      A software product implementing the algorithms of interpolation, classification, and optimal clustering is available from the authors.

77, TITLE:      Distributed Training of Embeddings using Graph Analytics
http://arxiv.org/abs/1909.03359
AUTHORS:        Gurbinder Gill ; Roshan Dathathri ; Saeed Maleki ; Madan Musuvathi ; Todd Mytkowicz ; Olli Saarikivi
HIGHLIGHT:      This paper presents a distributed training framework for a class of applications that use Skip-gram-like models to generate embeddings.

78, TITLE:      A Generative Framework for Zero-Shot Learning with Adversarial Domain Adaptation
http://arxiv.org/abs/1906.03038
AUTHORS:        Varun Khare ; Divyat Mahajan ; Homanga Bharadhwaj ; Vinay Verma ; Piyush Rai
COMMENTS:       Proceedings of Winter Conference on Applications of Computer Vision (WACV) 2020
HIGHLIGHT:      We present a domain adaptation based generative framework for zero-shot learning.

79, TITLE:      Discrete Infomax Codes for Supervised Representation Learning
http://arxiv.org/abs/1905.11656
AUTHORS:        Yoonho Lee ; Wonjae Kim ; Wonpyo Park ; Seungjin Choi
COMMENTS:       19 pages
HIGHLIGHT:      In this paper we present a model that produces Discrete InfoMax Codes (DIMCO); we learn a probabilistic encoder that yields k-way d-dimensional codes associated with input data.

80, TITLE:      Evolutionary reinforcement learning of dynamical large deviations
http://arxiv.org/abs/1909.00835
AUTHORS:        Stephen Whitelam ; Daniel Jacobson ; Isaac Tamblyn
HIGHLIGHT:      We show how to calculate the likelihood of dynamical large deviations using evolutionary reinforcement learning.

81, TITLE:      Nonparametric Continuous Sensor Registration
http://arxiv.org/abs/2001.04286
AUTHORS:        William Clark ; Maani Ghaffari ; Anthony Bloch
COMMENTS:       19 pages. arXiv admin note: substantial text overlap with arXiv:1904.02266
HIGHLIGHT:      This paper develops a new mathematical framework that enables nonparametric joint semantic/appearance and geometric representation of continuous functions using data.

82, TITLE:      Domain Knowledge Aided Explainable Artificial Intelligence for Intrusion Detection and Response
http://arxiv.org/abs/1911.09853
AUTHORS:        Sheikh Rabiul Islam ; William Eberle ; Sheikh K. Ghafoor ; Ambareen Siraj ; Mike Rogers
COMMENTS:       Accepted to be published in the Proceedings of the AAAI 2020 Spring Symposium on Combining Machine Learning and Knowledge Engineering in Practice (AAAI-MAKE 2020). Stanford University, Palo Alto, California, USA, March 23-25, 2020
HIGHLIGHT:      In this work, we infuse popular domain knowledge (i.e., CIA principles) in our model for better explainability and validate the approach on a network intrusion detection test case.

83, TITLE:      An ASP semantics for Constraints involving Conditional Aggregates
http://arxiv.org/abs/2002.06911
AUTHORS:        Pedro Cabalar ; Jorge Fandinno ; Torsten Schaub ; Philipp Wanko
HIGHLIGHT:      We elaborate upon the formal foundations of hybrid Answer Set Programming (ASP) and extend its underlying logical framework with aggregate functions over constraint values and variables.

84, TITLE:      BERTScore: Evaluating Text Generation with BERT
http://arxiv.org/abs/1904.09675
AUTHORS:        Tianyi Zhang ; Varsha Kishore ; Felix Wu ; Kilian Q. Weinberger ; Yoav Artzi
COMMENTS:       Code available at https://github.com/Tiiiger/bert_score; To appear in ICLR2020
HIGHLIGHT:      We propose BERTScore, an automatic evaluation metric for text generation.

85, TITLE:      Notes on Abstract Argumentation Theory
http://arxiv.org/abs/1806.07709
AUTHORS:        Anthony Peter Young
COMMENTS:       93 pages, 38 figures, 6 tables
HIGHLIGHT:      This note reviews Section 2 of Dung's seminal 1995 paper on abstract argumentation theory.

86, TITLE:      Towards Visually Explaining Variational Autoencoders
http://arxiv.org/abs/1911.07389
AUTHORS:        Wenqian Liu ; Runze Li ; Meng Zheng ; Srikrishna Karanam ; Ziyan Wu ; Bir Bhanu ; Richard J. Radke ; Octavia Camps
COMMENTS:       10 pages, 8 figures, 3 tables
HIGHLIGHT:      In this work, we take a step towards bridging this crucial gap, proposing the first technique to visually explain VAEs by means of gradient-based attention.

87, TITLE:      YOLOff: You Only Learn Offsets for robust 6DoF object pose estimation
http://arxiv.org/abs/2002.00911
AUTHORS:        Mathieu Gonzalez ; Amine Kacete ; Albert Murienne ; Eric Marchand
HIGHLIGHT:      In this paper, we propose a novel approach to perform 6 DoF object pose estimation from a single RGB-D image in cluttered scenes.

88, TITLE:      Improving the Harmony of the Composite Image by Spatial-Separated Attention Module
http://arxiv.org/abs/1907.06406
AUTHORS:        Xiaodong Cun ; Chi-Man Pun
COMMENTS:       Accepted by IEEE Transactions on Image Processing (TIP) 2020
HIGHLIGHT:      In this work, we start from an empirical observation: the differences can only be found in the spliced region between the spliced image and the harmonized result while they share the same semantic information and the appearance in the non-spliced region.

89, TITLE:      An Image Enhancing Pattern-based Sparsity for Real-time Inference on Mobile Devices
http://arxiv.org/abs/2001.07710
AUTHORS:        Xiaolong Ma ; Wei Niu ; Tianyun Zhang ; Sijia Liu ; Sheng Lin ; Hongjia Li ; Xiang Chen ; Jian Tang ; Kaisheng Ma ; Bin Ren ; Yanzhi Wang
COMMENTS:       arXiv admin note: text overlap with arXiv:1909.05073
HIGHLIGHT:      To solve the problem, we introduce a new sparsity dimension, namely pattern-based sparsity that comprises pattern and connectivity sparsity, and becoming both highly accurate and hardware friendly.

90, TITLE:      Topological Mapping for Manhattan-like Repetitive Environments
http://arxiv.org/abs/2002.06575
AUTHORS:        Sai Shubodh Puligilla ; Satyajit Tourani ; Tushar Vaidya ; Udit Singh Parihar ; Ravi Kiran Sarvadevabhatla ; K. Madhava Krishna
COMMENTS:       Second version with comparison with baseline SLAM system
HIGHLIGHT:      In the paper, we show that maintaining abstractions such as Topological Graph and Manhattan Graph help in recovering an accurate Pose Graph starting from a highly erroneous and unoptimized Pose Graph.

91, TITLE:      Deeply Activated Salient Region for Instance Search
http://arxiv.org/abs/2002.00185
AUTHORS:        Hui-Chu Xiao ; Wan-Lei Zhao ; Jie Lin ; Chong-Wah Ngo
COMMENTS:       10 pages, 8 figures
HIGHLIGHT:      In this paper, a simple but effective instance-level feature representation is presented.

92, TITLE:      Bipartite Perfect Matching as a Real Polynomial
http://arxiv.org/abs/2001.07642
AUTHORS:        Gal Beniamini ; Noam Nisan
HIGHLIGHT:      We obtain a description of the Bipartite Perfect Matching decision problem as a multilinear polynomial over the Reals.

93, TITLE:      Memory Augmented Recursive Neural Networks
http://arxiv.org/abs/1911.01545
AUTHORS:        Forough Arabshahi ; Zhichu Lu ; Sameer Singh ; Animashree Anandkumar
HIGHLIGHT:      To overcome this, we propose to augment each node in the recursive network with an external memory in the form of a differentiable stack.

94, TITLE:      Global Optimality Guarantees for Nonconvex Unsupervised Video Segmentation
http://arxiv.org/abs/1907.04409
AUTHORS:        Brendon G. Anderson ; Somayeh Sojoudi
COMMENTS:       Proceedings of the 57th Annual Allerton Conference on Communication, Control, and Computing, 2019; added funding source information and notation definitions
HIGHLIGHT:      In this paper, we consider the problem of unsupervised video object segmentation via background subtraction.

95, TITLE:      PAC-Bayes Control: Learning Policies that Provably Generalize to Novel Environments
http://arxiv.org/abs/1806.04225
AUTHORS:        Anirudha Majumdar ; Alec Farid ; Anoopkumar Sonar
COMMENTS:       Extended version of paper presented at the 2018 Conference on Robot Learning (CoRL)
HIGHLIGHT:      We propose policy learning algorithms that explicitly seek to minimize this upper bound.

96, TITLE:      Long-Range Indoor Navigation with PRM-RL
http://arxiv.org/abs/1902.09458
AUTHORS:        Anthony Francis ; Aleksandra Faust ; Hao-Tien Lewis Chiang ; Jasmine Hsu ; J. Chase Kew ; Marek Fiser ; Tsang-Wei Edward Lee
COMMENTS:       Accepted to T-RO
HIGHLIGHT:      We achieve this with PRM-RL, a hierarchical robot navigation method in which reinforcement learning agents that map noisy sensors to robot controls learn to solve short-range obstacle avoidance tasks, and then sampling-based planners map where these agents can reliably navigate in simulation; these roadmaps and agents are then deployed on robots, guiding them along the shortest path where the agents are likely to succeed.

97, TITLE:      IRIS: Implicit Reinforcement without Interaction at Scale for Learning Control from Offline Robot Manipulation Data
http://arxiv.org/abs/1911.05321
AUTHORS:        Ajay Mandlekar ; Fabio Ramos ; Byron Boots ; Silvio Savarese ; Li Fei-Fei ; Animesh Garg ; Dieter Fox
HIGHLIGHT:      In this paper, we propose Implicit Reinforcement without Interaction at Scale (IRIS), a novel framework for learning from large-scale demonstration datasets.

98, TITLE:      Compressed MRI Reconstruction Exploiting a Rotation-Invariant Total Variation Discretization
http://arxiv.org/abs/1911.11854
AUTHORS:        Erfan Ebrahim Esfahani ; Alireza Hosseini
HIGHLIGHT:      Inspired by the first-order method of Malitsky and Pock, we propose a novel variational framework for compressed MR image reconstruction which introduces the application of a rotation-invariant discretization of total variation functional into MR imaging while exploiting BM3D frame as a sparsifying transform.

99, TITLE:      Generator From Edges: Reconstruction of Facial Images
http://arxiv.org/abs/2002.06682
AUTHORS:        Nao Takano ; Gita Alaghband
HIGHLIGHT:      We propose Generator From Edges (GFE) [Figure 2].

100, TITLE:     A Multimodal Deep Network for the Reconstruction of T2W MR Images
http://arxiv.org/abs/1908.03009
AUTHORS:        Antonio Falvo ; Danilo Comminiello ; Simone Scardapane ; Michele Scarpiniti ; Aurelio Uncini
COMMENTS:       29th Italian Neural Networks Workshop (WIRN 2019)
HIGHLIGHT:      In this paper, we present a deep learning method that is able to reconstruct subsampled MR images obtained by reducing the k-space data, while maintaining a high image quality that can be used to observe brain lesions.