2020.03.20.txt

==========New Papers==========
1, TITLE:       Pose Augmentation: Class-agnostic Object Pose Transformation for Object Recognition
http://arxiv.org/abs/2003.08526
AUTHORS:        Yunhao Ge ; Jiaping Zhao ; Laurent Itti
COMMENTS:       21 pages(including references) 6 figures
HIGHLIGHT:      Here, we propose a different approach: a class-agnostic object pose transformation network (OPT-Net) can transform an image along 3D yaw and pitch axes to synthesize additional poses continuously.

2, TITLE:       Detecting Lane and Road Markings at A Distance with Perspective Transformer Layers
http://arxiv.org/abs/2003.08550
AUTHORS:        Zhuoping Yu ; Xiaozhou Ren ; Yuyao Huang ; Wei Tian ; Junqiao Zhao
HIGHLIGHT:      To solve this problem, we adopt the Encoder-Decoder architecture in Fully Convolutional Networks and leverage the idea of Spatial Transformer Networks to introduce a novel semantic segmentation neural network.

3, TITLE:       Joint Event Extraction along Shortest Dependency Paths using Graph Convolutional Networks
http://arxiv.org/abs/2003.08615
AUTHORS:        Ali Balali ; Masoud Asadpour ; Ricardo Campos ; Adam Jatowt
HIGHLIGHT:      To address the two above-referred problems, we propose a novel joint event extraction framework that aims to extract multiple event triggers and arguments simultaneously by introducing shortest dependency path (SDP) in the dependency graph.

4, TITLE:       Efficient and Robust Shape Correspondence via Sparsity-Enforced Quadratic Assignment
http://arxiv.org/abs/2003.08680
AUTHORS:        Rui Xiang ; Rongjie Lai ; Hongkai Zhao
COMMENTS:       8 pages, 6 figures, to be published in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
HIGHLIGHT:      We use various experiments to show the efficiency, quality, and versatility of our method on large data sets, patches, and point clouds (without global meshes).

5, TITLE:       Detecting Deepfakes with Metric Learning
http://arxiv.org/abs/2003.08645
AUTHORS:        Akash Kumar ; Arnav Bhavsar
HIGHLIGHT:      In this work, we analyze several deep learning approaches in the context of deepfakes classification in high compression scenario and demonstrate that a proposed approach based on metric learning can be very effective in performing such a classification.

6, TITLE:       Unsupervised text line segmentation
http://arxiv.org/abs/2003.08632
AUTHORS:        Berat Kurar Barakat ; Ahmad Droby ; Rym Alasam ; Boraq Madi ; Irina Rabaev ; Raed Shammes ; Jihad El-Sana
COMMENTS:       First attempt that uses unsupervised deep CNN for text line segmentation
HIGHLIGHT:      We present an unsupervised text line segmentation method that is inspired by the relative variance between text lines and spaces among text lines.

7, TITLE:       Photo-Realistic Video Prediction on Natural Videos of Largely Changing Frames
http://arxiv.org/abs/2003.08635
AUTHORS:        Osamu Shouno
COMMENTS:       16 pages, 7 figures
HIGHLIGHT:      To address these issues, we propose a deep residual network with the hierarchical architecture where each layer makes a prediction of future state at different spatial resolution, and these predictions of different layers are merged via top-down connections to generate future frames.

8, TITLE:       LANCE: efficient low-precision quantized Winograd convolution for neural networks based on graphics processing units
http://arxiv.org/abs/2003.08646
AUTHORS:        Guangli Li ; Lei Liu ; Xueying Wang ; Xiu Ma ; Xiaobing Feng
COMMENTS:       Accepted by ICASSP 2020
HIGHLIGHT:      In this paper, we propose an efficient low-precision quantized Winograd convolution algorithm, called LANCE, which combines the advantages of fast convolution and quantization techniques.

9, TITLE:       Domain-Adaptive Few-Shot Learning
http://arxiv.org/abs/2003.08626
AUTHORS:        An Zhao ; Mingyu Ding ; Zhiwu Lu ; Tao Xiang ; Yulei Niu ; Jiechao Guan ; Ji-Rong Wen ; Ping Luo
HIGHLIGHT:      To this end, we propose a novel domain-adversarial prototypical network (DAPN) model.

10, TITLE:      Depth Potentiality-Aware Gated Attention Network for RGB-D Salient Object Detection
http://arxiv.org/abs/2003.08608
AUTHORS:        Zuyao Chen ; Qingming Huang
HIGHLIGHT:      In this paper, we address these two issues in a holistic model synergistically, and propose a novel network named DPANet to explicitly model the potentiality of the depth map and effectively integrate the cross-modal complementarity.

11, TITLE:      PT2PC: Learning to Generate 3D Point Cloud Shapes from Part Tree Conditions
http://arxiv.org/abs/2003.08624
AUTHORS:        Kaichun Mo ; He Wang ; Xinchen Yan ; Leonidas J. Guibas
HIGHLIGHT:      In order to learn such a conditional shape generation procedure in an end-to-end fashion, we propose a conditional GAN "part tree"-to-"point cloud" model (PT2PC) that disentangles the structural and geometric factors.

12, TITLE:      AQPDBJUT Dataset: Picture-Based PM2.5 Monitoring in the Campus of BJUT
http://arxiv.org/abs/2003.08609
AUTHORS:        Yonghui Zhang ; Ke Gu ; Zhifang Xia ; Junfei Qiao
HIGHLIGHT:      As the source of PM prevention and control, developing a good model for PM monitoring is extremely urgent and has posed a big challenge.

13, TITLE:      Foldover Features for Dynamic Object Behavior Description in Microscopic Videos
http://arxiv.org/abs/2003.08628
AUTHORS:        Xialin Li ; Chen Li ; Wenwei Zhao
HIGHLIGHT:      To this end, we propose foldover features to describe the behavior of dynamic objects.

14, TITLE:      A Corpus of Adpositional Supersenses for Mandarin Chinese
http://arxiv.org/abs/2003.08437
AUTHORS:        Siyao Peng ; Yang Liu ; Yilun Zhu ; Austin Blodgett ; Yushi Zhao ; Nathan Schneider
COMMENTS:       LREC 2020 camera-ready
HIGHLIGHT:      This paper presents a corpus in which all adpositions have been semantically annotated in Mandarin Chinese; to the best of our knowledge, this is the first Chinese corpus to be broadly annotated with adposition semantics.

15, TITLE:      DRST: Deep Residual Shearlet Transform for Densely Sampled Light Field Reconstruction
http://arxiv.org/abs/2003.08865
AUTHORS:        Yuan Gao ; Robert Bregovic ; Reinhard Koch ; Atanas Gotchev
HIGHLIGHT:      To overcome this limitation, a novel learning-based ST approach, referred to as Deep Residual Shearlet Transform (DRST), is proposed in this paper.

16, TITLE:      GAN Compression: Efficient Architectures for Interactive Conditional GANs
http://arxiv.org/abs/2003.08936
AUTHORS:        Muyang Li ; Ji Lin ; Yaoyao Ding ; Zhijian Liu ; Jun-Yan Zhu ; Song Han
HIGHLIGHT:      In this work, we propose a general-purpose compression framework for reducing the inference time and model size of the generator in cGANs.

17, TITLE:      NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
http://arxiv.org/abs/2003.08934
AUTHORS:        Ben Mildenhall ; Pratul P. Srinivasan ; Matthew Tancik ; Jonathan T. Barron ; Ravi Ramamoorthi ; Ren Ng
COMMENTS:       Project page with videos and code: http://tancik.com/nerf
HIGHLIGHT:      We present a method that achieves state-of-the-art results for synthesizing novel views of complex scenes by optimizing an underlying continuous volumetric scene function using a sparse set of input views.

18, TITLE:      Group Sparsity: The Hinge Between Filter Pruning and Decomposition for Network Compression
http://arxiv.org/abs/2003.08935
AUTHORS:        Yawei Li ; Shuhang Gu ; Christoph Mayer ; Luc Van Gool ; Radu Timofte
COMMENTS:       Accepted by CVPR2020. Code is available at https://github.com/ofsoundof/group_sparsity
HIGHLIGHT:      In this paper, we analyze two popular network compression techniques, i.e. filter pruning and low-rank decomposition, in a unified sense.

19, TITLE:      Overinterpretation reveals image classification model pathologies
http://arxiv.org/abs/2003.08907
AUTHORS:        Brandon Carter ; Siddhartha Jain ; Jonas Mueller ; David Gifford
HIGHLIGHT:      We find that high scoring convolutional neural networks (CNN) exhibit troubling pathologies that allow them to display high accuracy even in the absence of semantically salient features.

20, TITLE:      Efficiently Calibrating Cable-Driven Surgical Robots With RGBD Sensing, Temporal Windowing, and Linear and Recurrent Neural Network Compensation
http://arxiv.org/abs/2003.08520
AUTHORS:        Minho Hwang ; Brijen Thananjeyan ; Samuel Paradis ; Daniel Seita ; Jeffrey Ichnowski ; Danyal Fer ; Thomas Low ; Ken Goldberg
COMMENTS:       9 pages, 11 figures, 2 tables
HIGHLIGHT:      We propose a novel approach to efficiently calibrate a dVRK by placing a 3D printed fiducial coordinate frame on the arm and end-effector that is tracked using RGBD sensing.

21, TITLE:      Local Rotation Invariance in 3D CNNs
http://arxiv.org/abs/2003.08890
AUTHORS:        Vincent Andrearczyk ; Julien Fageot ; Valentin Oreiller ; Xavier Montet ; Adrien Depeursinge
HIGHLIGHT:      In this paper, we propose and compare several methods to obtain LRI CNNs with directional sensitivity.

22, TITLE:      Across Scales \& Across Dimensions: Temporal Super-Resolution using Deep Internal Learning
http://arxiv.org/abs/2003.08872
AUTHORS:        Liad Pollak Zuckerman ; Shai Bagon ; Eyal Naor ; George Pisha ; Michal Irani
HIGHLIGHT:      In this paper we propose a "Deep Internal Learning" approach for true TSR.

23, TITLE:      Depth Estimation by Learning Triangulation and Densification of Sparse Points for Multi-view Stereo
http://arxiv.org/abs/2003.08933
AUTHORS:        Ayan Sinha ; Zak Murez ; James Bartolozzi ; Vijay Badrinarayanan ; Andrew Rabinovich
HIGHLIGHT:      Distinct from cost volume approaches, we propose an efficient depth estimation approach by first (a) detecting and evaluating descriptors for interest points, then (b) learning to match and triangulate a small set of interest points, and finally (c) densifying this sparse set of 3D points using CNNs.

24, TITLE:      Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation
http://arxiv.org/abs/2003.08866
AUTHORS:        Zhenda Xie ; Zheng Zhang ; Xizhou Zhu ; Gao Huang ; Stephen Lin
HIGHLIGHT:      Towards reducing this superfluous computation, we propose to compute features only at sparsely sampled locations, which are probabilistically chosen according to activation responses, and then densely reconstruct the feature map with an efficient interpolation procedure.

25, TITLE:      Normalized and Geometry-Aware Self-Attention Network for Image Captioning
http://arxiv.org/abs/2003.08897
AUTHORS:        Longteng Guo ; Jing Liu ; Xinxin Zhu ; Peng Yao ; Shichen Lu ; Hanqing Lu
COMMENTS:       Accepted by CVPR 2020
HIGHLIGHT:      In this paper, we improve SA from two aspects to promote the performance of image captioning.

26, TITLE:      Unique Geometry and Texture from Corresponding Image Patches
http://arxiv.org/abs/2003.08885
AUTHORS:        Dor Verbin ; Steven J. Gortler ; Todd Zickler
HIGHLIGHT:      We present a sufficient condition for the recovery of a unique texture process and a unique set of viewpoints from a set of image patches that are generated by observing a flat texture process from unknown directions and orientations.

27, TITLE:      Enhanced POET: Open-Ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions
http://arxiv.org/abs/2003.08536
AUTHORS:        Rui Wang ; Joel Lehman ; Aditya Rawal ; Jiale Zhi ; Yulun Li ; Jeff Clune ; Kenneth O. Stanley
COMMENTS:       23 pages, 14 figures
HIGHLIGHT:      Here we introduce and empirically validate two new innovations to the original algorithm, as well as two external innovations designed to help elucidate its full potential.

28, TITLE:      Goal-Conditioned End-to-End Visuomotor Control for Versatile Skill Primitives
http://arxiv.org/abs/2003.08854
AUTHORS:        Oliver Groth ; Chia-Man Hung ; Andrea Vedaldi ; Ingmar Posner
COMMENTS:       15 pages, 10 figures, 9 tables; supplementary video available: https://youtu.be/zn_lPor9zCU
HIGHLIGHT:      In this paper we propose a conditioning scheme which avoids these pitfalls by learning the controller and its conditioning in an end-to-end manner.

29, TITLE:      Learning to Fly via Deep Model-Based Reinforcement Learning
http://arxiv.org/abs/2003.08876
AUTHORS:        Philip Becker-Ehmck ; Maximilian Karl ; Jan Peters ; Patrick van der Smagt
HIGHLIGHT:      In this work, by leveraging a learnt probabilistic model of drone dynamics, we achieve human-like quadrotor control through model-based reinforcement learning.

30, TITLE:      CPR-GCN: Conditional Partial-Residual Graph Convolutional Network in Automated Anatomical Labeling of Coronary Arteries
http://arxiv.org/abs/2003.08560
AUTHORS:        Han Yang ; Xingjian Zhen ; Ying Chi ; Lei Zhang ; Xian-Sheng Hua
COMMENTS:       This work is done by Xingjian Zhen during internship in Alibaba Damo Academia
HIGHLIGHT:      Motivated by the wide application of the graph neural network in structured data, in this paper, we propose a conditional partial-residual graph convolutional network (CPR-GCN), which takes both position and CT image into consideration, since CT image contains abundant information such as branch size and spanning direction.

31, TITLE:      Unsupervised Domain Adaptation via Structurally Regularized Deep Clustering
http://arxiv.org/abs/2003.08607
AUTHORS:        Hui Tang ; Ke Chen ; Kui Jia
COMMENTS:       14 pages, 5 figures, 8 tables, accepted by CVPR2020 - Oral
HIGHLIGHT:      To alleviate this risk, we are motivated by the assumption of structural domain similarity, and propose to directly uncover the intrinsic target discrimination via discriminative clustering of target data.

32, TITLE:      Quality Control of Neuron Reconstruction Based on Deep Learning
http://arxiv.org/abs/2003.08556
AUTHORS:        Donghuan Lu ; Sujun Zhao ; Peng Xie ; Kai Ma ; Lijuan Liu ; Yefeng Zheng
COMMENTS:       9 pages, 2 figures
HIGHLIGHT:      To ensure the quality of reconstructed neurons and provide guidance for annotators to improve their efficiency, we propose a deep learning based quality control method for neuron reconstruction in this paper.

33, TITLE:      High Accuracy Face Geometry Capture using a Smartphone Video
http://arxiv.org/abs/2003.08583
AUTHORS:        Shubham Agrawal ; Anuj Pahuja ; Simon Lucey
COMMENTS:       Presented at The IEEE Winter Conference on Applications of Computer Vision (WACV), 2020, pp. 81-90
HIGHLIGHT:      We attempt to answer this question in our work.

34, TITLE:      Curriculum DeepSDF
http://arxiv.org/abs/2003.08593
AUTHORS:        Yueqi Duan ; Haidong Zhu ; He Wang ; Li Yi ; Ram Nevatia ; Leonidas J. Guibas
HIGHLIGHT:      In this paper, we design a "shape curriculum" for learning continuous Signed Distance Function (SDF) on shapes, namely Curriculum DeepSDF.

35, TITLE:      Deep convolutional embedding for digitized painting clustering
http://arxiv.org/abs/2003.08597
AUTHORS:        Giovanna Castellano ; Gennaro Vessio
HIGHLIGHT:      To address these issues, we propose a deep convolutional embedding model for clustering digital paintings, in which the task of mapping the input raw data to an abstract, latent space is optimized jointly with the task of finding a set of cluster centroids in this latent feature space.

36, TITLE:      QnAMaker: Data to Bot in 2 Minutes
http://arxiv.org/abs/2003.08553
AUTHORS:        Parag Agrawal ; Tulasi Menon ; Aya Kamel ; Michel Naim ; Chaikesh Chouragade ; Gurvinder Singh ; Rohan Kulkarni ; Anshuman Suri ; Sahithi Katakam ; Vineet Pratik ; Prakul Bansal ; Simerpreet Kaur ; Neha Rajput ; Anand Duggal ; Achraf Chalabi ; Prashant Choudhari ; Reddy Satti ; Niranjan Nayak
COMMENTS:       Published at The Web Conference 2020 in the demo track
HIGHLIGHT:      We demonstrate QnAMaker, a service that creates a conversational layer over semi-structured data such as FAQ pages, product manuals, and support documents.

37, TITLE:      Boosting Factual Correctness of Abstractive Summarization with Knowledge Graph
http://arxiv.org/abs/2003.08612
AUTHORS:        Chenguang Zhu ; William Hinthorn ; Ruochen Xu ; Qingkai Zeng ; Michael Zeng ; Xuedong Huang ; Meng Jiang
COMMENTS:       14 pages, 2 figures
HIGHLIGHT:      In this paper, we propose to boost factual correctness of summaries via the fusion of knowledge, i.e. extracted factual relations from the article.

38, TITLE:      Diversity, Density, and Homogeneity: Quantitative Characteristic Metrics for Text Collections
http://arxiv.org/abs/2003.08529
AUTHORS:        Yi-An Lai ; Xuan Zhu ; Yi Zhang ; Mona Diab
COMMENTS:       Accepted by LREC 2020
HIGHLIGHT:      In this work, we propose metrics of diversity, density, and homogeneity that quantitatively measure the dispersion, sparsity, and uniformity of a text collection.

39, TITLE:      Beheshti-NER: Persian Named Entity Recognition Using BERT
http://arxiv.org/abs/2003.08875
AUTHORS:        Ehsan Taher ; Seyed Abbas Hoseini ; Mehrnoush Shamsfard
HIGHLIGHT:      In this paper, we use the pre-trained deep bidirectional network, BERT, to make a model for named entity recognition in Persian.

40, TITLE:      An Analysis on the Learning Rules of the Skip-Gram Model
http://arxiv.org/abs/2003.08489
AUTHORS:        Canlin Zhang ; Xiuwen Liu ; Daniel Bis
COMMENTS:       Published on the 2019 International Joint Conference on Neural Networks
HIGHLIGHT:      In this work, we derive the learning rules for the skip-gram model and establish their close relationship to competitive learning.

41, TITLE:      Collaborative Distillation for Ultra-Resolution Universal Style Transfer
http://arxiv.org/abs/2003.08436
AUTHORS:        Huan Wang ; Yijun Li ; Yuehai Wang ; Haoji Hu ; Ming-Hsuan Yang
COMMENTS:       Accepted by CVPR 2020
HIGHLIGHT:      In this work, we present a new knowledge distillation method (named Collaborative Distillation) for encoder-decoder based neural style transfer to reduce the convolutional filters.

42, TITLE:      STEm-Seg: Spatio-temporal Embeddings for Instance Segmentation in Videos
http://arxiv.org/abs/2003.08429
AUTHORS:        Ali Athar ; Sabarinath Mahadevan ; Aljoša Ošep ; Laura Leal-Taixé ; Bastian Leibe
COMMENTS:       28 pages, 6 figures
HIGHLIGHT:      In this paper, we propose a different approach that is well-suited to a variety of tasks involving instance segmentation in videos.

43, TITLE:      Adversarial Texture Optimization from RGB-D Scans
http://arxiv.org/abs/2003.08400
AUTHORS:        Jingwei Huang ; Justus Thies ; Angela Dai ; Abhijit Kundu ; Chiyu Max Jiang ; Leonidas Guibas ; Matthias Nießner ; Thomas Funkhouser
HIGHLIGHT:      In this work, we present a novel approach for color texture generation using a conditional adversarial loss obtained from weakly-supervised views.

44, TITLE:      A Content Transformation Block For Image Style Transfer
http://arxiv.org/abs/2003.08407
AUTHORS:        Dmytro Kotovenko ; Artsiom Sanakoyeu ; Pingchuan Ma ; Sabine Lang ; Björn Ommer
COMMENTS:       Accepted to CVPR 2019
HIGHLIGHT:      Therefore, we introduce a content transformation module between the encoder and decoder.

45, TITLE:      Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation
http://arxiv.org/abs/2003.08440
AUTHORS:        Yingda Xia ; Yi Zhang ; Fengze Liu ; Wei Shen ; Alan Yuille
COMMENTS:       The first two authors contributed equally to this work
HIGHLIGHT:      In this paper, we systematically study failure and anomaly detection for semantic segmentation and propose a unified framework, consisting of two modules, to address these two related problems.

46, TITLE:      Lifelong Learning with Searchable Extension Units
http://arxiv.org/abs/2003.08559
AUTHORS:        Wenjin Wang ; Yunqing Hu ; Yin Zhang
HIGHLIGHT:      To solve those problems, in this paper, we propose a new lifelong learning framework named Searchable Extension Units (SEU) by introducing Neural Architecture Search into lifelong learning, which breaks down the need for a predefined original model and searches for specific extension units for different tasks, without compromising the performance of the model on different tasks.

47, TITLE:      Ensemble learning in CNN augmented with fully connected subnetworks
http://arxiv.org/abs/2003.08562
AUTHORS:        Daiki Hirata ; Norikazu Takahashi
COMMENTS:       6 pages, 2 figures, 5 tables
HIGHLIGHT:      In this paper, we propose a new model called EnsNet which is composed of one base CNN and multiple Fully Connected SubNetworks (FCSNs).

48, TITLE:      Decentralized MCTS via Learned Teammate Models
http://arxiv.org/abs/2003.08727
AUTHORS:        Aleksander Czechowski ; Frans Oliehoek
HIGHLIGHT:      In this paper we present a policy improvement operator for learning to plan in iterated cooperative multi-agent scenarios.

49, TITLE:      Utilizing Language Relatedness to improve Machine Translation: A Case Study on Languages of the Indian Subcontinent
http://arxiv.org/abs/2003.08925
AUTHORS:        Anoop Kunchukuttan ; Pushpak Bhattacharyya
COMMENTS:       This work was done in 2017-2018 as part of the first author's thesis
HIGHLIGHT:      In this work, we present an extensive study of statistical machine translation involving languages of the Indian subcontinent.

50, TITLE:      Train Scheduling with Hybrid Answer Set Programming
http://arxiv.org/abs/2003.08598
AUTHORS:        Dirk Abels ; Julian Jordi ; Max Ostrowski ; Torsten Schaub ; Ambra Toletti ; Philipp Wanko
COMMENTS:       Under consideration in Theory and Practice of Logic Programming (TPLP)
HIGHLIGHT:      We present a solution to real-world train scheduling problems, involving routing, scheduling, and optimization, based on Answer Set Programming (ASP).

51, TITLE:      Adjust Planning Strategies to Accommodate Reinforcement Learning Agents
http://arxiv.org/abs/2003.08554
AUTHORS:        Xuerun Chen
HIGHLIGHT:      Two methods focus on micro and macro action respectively.

52, TITLE:      Placement Optimization with Deep Reinforcement Learning
http://arxiv.org/abs/2003.08445
AUTHORS:        Anna Goldie ; Azalia Mirhoseini
COMMENTS:       International Symposium on Physical Design (ISPD), 2020
HIGHLIGHT:      In this paper, we start by motivating reinforcement learning as a solution to the placement problem.

53, TITLE:      Semi-supervised few-shot learning for medical image segmentation
http://arxiv.org/abs/2003.08462
AUTHORS:        Abdur R Feyjie ; Reza Azad ; Marco Pedersoli ; Claude Kauffman ; Ismail Ben Ayed ; Jose Dolz
HIGHLIGHT:      In this work, we propose a novel few-shot learning framework for semantic segmentation, where unlabeled images are also made available at each episode.

54, TITLE:      SAPIEN: A SimulAted Part-based Interactive ENvironment
http://arxiv.org/abs/2003.08515
AUTHORS:        Fanbo Xiang ; Yuzhe Qin ; Kaichun Mo ; Yikuan Xia ; Hao Zhu ; Fangchen Liu ; Minghua Liu ; Hanxiao Jiang ; Yifu Yuan ; He Wang ; Li Yi ; Angel X. Chang ; Leonidas J. Guibas ; Hao Su
HIGHLIGHT:      Our SAPIEN enables various robotic vision and interaction tasks that require detailed part-level understanding.We evaluate state-of-the-art vision algorithms for part detection and motion attribute recognition as well as demonstrate robotic interaction tasks using heuristic approaches and reinforcement learning algorithms.

55, TITLE:      Evaluating Salient Object Detection in Natural Images with Multiple Objects having Multi-level Saliency
http://arxiv.org/abs/2003.08514
AUTHORS:        Gökhan Yildirim ; Debashis Sen ; Mohan Kankanhalli ; Sabine Süsstrunk
COMMENTS:       Accepted Article
HIGHLIGHT:      In this paper, we corroborate based on three subjective experiments on a novel image dataset that objects in natural images are inherently perceived to have varying levels of importance.

56, TITLE:      Self-Supervised Contextual Bandits in Computer Vision
http://arxiv.org/abs/2003.08485
AUTHORS:        Aniket Anand Deshmukh ; Abhimanu Kumar ; Levi Boyles ; Denis Charles ; Eren Manavoglu ; Urun Dogan
HIGHLIGHT:      We provide a novel approach to tackle this issue by combining a contextual bandit objective with a self supervision objective.

57, TITLE:      Gaze-Sensing LEDs for Head Mounted Displays
http://arxiv.org/abs/2003.08499
AUTHORS:        Kaan Akşit ; Jan Kautz ; David Luebke
COMMENTS:       14 pages, 7 figures. THIS WORK WAS CONDUCTED IN 2015
HIGHLIGHT:      We introduce a new gaze tracker for Head Mounted Displays (HMDs).

58, TITLE:      Visual link retrieval and knowledge discovery in painting datasets
http://arxiv.org/abs/2003.08476
AUTHORS:        Giovanna Castellano ; Eufemia Lella ; Gennaro Vessio
COMMENTS:       submitted to Multimedia Tools and Applications
HIGHLIGHT:      To help art historians better understand visual arts, the present paper presents a framework for visual link retrieval and knowledge discovery in digital painting datasets.

59, TITLE:      Reconstructing Sinus Anatomy from Endoscopic Video -- Towards a Radiation-free Approach for Quantitative Longitudinal Assessment
http://arxiv.org/abs/2003.08502
AUTHORS:        Xingtong Liu ; Maia Stiber ; Jindan Huang ; Masaru Ishii ; Gregory D. Hager ; Russell H. Taylor ; Mathias Unberath
HIGHLIGHT:      We present a patient-specific, learning-based method for 3D reconstruction of sinus surface anatomy directly and only from endoscopic videos.

60, TITLE:      A Metric Learning Reality Check
http://arxiv.org/abs/2003.08505
AUTHORS:        Kevin Musgrave ; Serge Belongie ; Ser-Nam Lim
HIGHLIGHT:      In this paper, we take a closer look at the field to see if this is actually true.

61, TITLE:      MINT: Deep Network Compression via Mutual Information-based Neuron Trimming
http://arxiv.org/abs/2003.08472
AUTHORS:        Madan Ravi Ganesh ; Jason J. Corso ; Salimeh Yasaei Sekeh
COMMENTS:       12 pages
HIGHLIGHT:      Most approaches to deep neural network compression via pruning either evaluate a filter's importance using its weights or optimize an alternative objective function with sparsity constraints.

62, TITLE:      Train, Learn, Expand, Repeat
http://arxiv.org/abs/2003.08469
AUTHORS:        Abhijeet Parida ; Aadhithya Sankar ; Rami Eisawy ; Tom Finck ; Benedikt Wiestler ; Franz Pfister ; Julia Moosbauer
COMMENTS:       Published as a workshop paper at AI4AH, ICLR 2020
HIGHLIGHT:      We propose a recursive training strategy to perform the task of semantic segmentation given only very few training samples with pixel-level annotations.

63, TITLE:      HyNNA: Improved Performance for Neuromorphic Vision Sensor based Surveillance using Hybrid Neural Network Architecture
http://arxiv.org/abs/2003.08603
AUTHORS:        Deepak Singla ; Soham Chatterjee ; Lavanya Ramapantulu ; Andres Ussa ; Bharath Ramesh ; Arindam Basu
COMMENTS:       4 pages, 2 figures
HIGHLIGHT:      To address this, we improve on a recently proposed hybrid event-frame approach by using morphological image processing algorithms for region proposal and address the low-power requirement for object detection and classification by exploring various convolutional neural network (CNN) architectures.

64, TITLE:      RADIOGAN: Deep Convolutional Conditional Generative adversarial Network To Generate PET Images
http://arxiv.org/abs/2003.08663
AUTHORS:        Amine Amyar ; Su Ruan ; Pierre Vera ; Pierre Decazes ; Romain Modzelewski
COMMENTS:       4 pages, 5 figures
HIGHLIGHT:      In this paper, we propose a deep convolutional conditional generative adversarial network to generate MIP positron emission tomography image (PET) which is a 2D image that represents a 3D volume for fast interpretation, according to different lesions or non lesion (normal).

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

66, TITLE:      Automatic accuracy management of quantum programs via (near-)symbolic resource estimation
http://arxiv.org/abs/2003.08408
AUTHORS:        Giulia Meuli ; Mathias Soeken ; Martin Roetteler ; Thomas Häner
COMMENTS:       15 pages
HIGHLIGHT:      We propose a methodology that tracks such errors automatically and solves the optimization problem of finding accuracy parameters that guarantee a specified overall accuracy while aiming to minimize a custom implementation cost.

67, TITLE:      Pedestrian Detection: The Elephant In The Room
http://arxiv.org/abs/2003.08799
AUTHORS:        Irtiza Hasan ; Shengcai Liao ; Jinpeng Li ; Saad Ullah Akram ; Ling Shao
COMMENTS:       17 pages, 1 figure
HIGHLIGHT:      To this end, we conduct a comprehensive study in this paper, using a general principle of direct cross-dataset evaluation.

68, TITLE:      Exemplar Normalization for Learning Deep Representation
http://arxiv.org/abs/2003.08761
AUTHORS:        Ruimao Zhang ; Zhanglin Peng ; Lingyun Wu ; Zhen Li ; Ping Luo
COMMENTS:       Accepted by CVPR2020, normalization methods, image classification
HIGHLIGHT:      This work investigates a novel dynamic learning-to-normalize (L2N) problem by proposing Exemplar Normalization (EN), which is able to learn different normalization methods for different convolutional layers and image samples of a deep network.

69, TITLE:      Leveraging Frequency Analysis for Deep Fake Image Recognition
http://arxiv.org/abs/2003.08685
AUTHORS:        Joel Frank ; Thorsten Eisenhofer ; Lea Schönherr ; Asja Fischer ; Dorothea Kolossa ; Thorsten Holz
COMMENTS:       For accompanying code see https://github.com/RUB-SysSec/GANDCTAnalysis. 13 pages, 7 figures
HIGHLIGHT:      In this paper, we address this shortcoming and our results reveal that in frequency space, GAN-generated images exhibit severe artifacts that can be easily identified.

70, TITLE:      Multi-task Collaborative Network for Joint Referring Expression Comprehension and Segmentation
http://arxiv.org/abs/2003.08813
AUTHORS:        Gen Luo ; Yiyi Zhou ; Xiaoshuai Sun ; Liujuan Cao ; Chenglin Wu ; Cheng Deng ; Rongrong Ji
COMMENTS:       accpected by CVPR2020 (oral)
HIGHLIGHT:      In this paper, we propose a novel Multi-task Collaborative Network (MCN) to achieve a joint learning of REC and RES for the first time.

71, TITLE:      Self-Guided Adaptation: Progressive Representation Alignment for Domain Adaptive Object Detection
http://arxiv.org/abs/2003.08777
AUTHORS:        Zongxian Li ; Qixiang Ye ; Chong Zhang ; Jingjing Liu ; Shijian Lu ; Yonghong Tian
HIGHLIGHT:      In this work, we propose a Self-Guided Adaptation (SGA) model, target at aligning feature representation and transferring object detection models across domains while considering the instantaneous alignment difficulty.

72, TITLE:      Giving Commands to a Self-driving Car: A Multimodal Reasoner for Visual Grounding
http://arxiv.org/abs/2003.08717
AUTHORS:        Thierry Deruyttere ; Guillem Collell ; Marie-Francine Moens
COMMENTS:       14 pages + 22 pages supplementary with a lot of figures
HIGHLIGHT:      We propose a new spatial memory module and a spatial reasoner for the Visual Grounding (VG) task.

73, TITLE:      High-Resolution Daytime Translation Without Domain Labels
http://arxiv.org/abs/2003.08791
AUTHORS:        Ivan Anokhin ; Pavel Solovev ; Denis Korzhenkov ; Alexey Kharlamov ; Taras Khakhulin ; Gleb Sterkin ; Alexey Silvestrov ; Sergey Nikolenko ; Victor Lempitsky
COMMENTS:       accepted to CVPR 2020
HIGHLIGHT:      We present the high-resolution daytime translation (HiDT) model for this task.

74, TITLE:      Backdooring and Poisoning Neural Networks with Image-Scaling Attacks
http://arxiv.org/abs/2003.08633
AUTHORS:        Erwin Quiring ; Konrad Rieck
COMMENTS:       IEEE Deep Learning and Security Workshop (DLS) 2020
HIGHLIGHT:      In this paper, we propose a novel strategy for hiding backdoor and poisoning attacks.

75, TITLE:      Extremal Region Analysis based Deep Learning Framework for Detecting Defects
http://arxiv.org/abs/2003.08525
AUTHORS:        Zelin Deng ; Xiaolong Yan ; Shengjun Zhang ; Colleen P. Bailey
HIGHLIGHT:      A maximally stable extreme region (MSER) analysis based convolutional neural network (CNN) for unified defect detection framework is proposed in this paper.



==========Updates to Previous Papers==========
1, TITLE:       Heterogeneity Loss to Handle Intersubject and Intrasubject Variability in Cancer
http://arxiv.org/abs/2003.03295
AUTHORS:        Shubham Goswami ; Suril Mehta ; Dhruva Sahrawat ; Anubha Gupta ; Ritu Gupta
COMMENTS:       Accepted in ICLR 2020 workshop AI4AH(https://sites.google.com/view/ai4ah-iclr2020)
HIGHLIGHT:      In this work, we address these problems in the application of B-cell Acute Lymphoblastic Leukemia (B-ALL) diagnosis using deep learning.

2, TITLE:       High-Order Information Matters: Learning Relation and Topology for Occluded Person Re-Identification
http://arxiv.org/abs/2003.08177
AUTHORS:        Guan'an Wang ; Shuo Yang ; Huanyu Liu ; Zhicheng Wang ; Yang Yang ; Shuliang Wang ; Gang Yu ; Erjin Zhou ; Jian Sun
COMMENTS:       accepted by CVPR'20
HIGHLIGHT:      In this paper, we propose a novel framework by learning high-order relation and topology information for discriminative features and robust alignment.

3, TITLE:       Perceptual Image Super-Resolution with Progressive Adversarial Network
http://arxiv.org/abs/2003.03756
AUTHORS:        Lone Wong ; Deli Zhao ; Shaohua Wan ; Bo Zhang
HIGHLIGHT:      In this paper, we argue that the curse of dimensionality is the underlying reason of limiting the performance of state-of-the-art algorithms.

4, TITLE:       The Group Loss for Deep Metric Learning
http://arxiv.org/abs/1912.00385
AUTHORS:        Ismail Elezi ; Sebastiano Vascon ; Alessandro Torcinovich ; Marcello Pelillo ; Laura Leal-Taixe
HIGHLIGHT:      We propose Group Loss, a loss function based on a differentiable label-propagation method that enforces embedding similarity across all samples of a group while promoting, at the same time, low-density regions amongst data points belonging to different groups.

5, TITLE:       GraphTER: Unsupervised Learning of Graph Transformation Equivariant Representations via Auto-Encoding Node-wise Transformations
http://arxiv.org/abs/1911.08142
AUTHORS:        Xiang Gao ; Wei Hu ; Guo-Jun Qi
HIGHLIGHT:      To this end, we propose a novel unsupervised learning of Graph Transformation Equivariant Representations (GraphTER), aiming to capture intrinsic patterns of graph structure under both global and local transformations.

6, TITLE:       Face Anti-Spoofing by Learning Polarization Cues in a Real-World Scenario
http://arxiv.org/abs/2003.08024
AUTHORS:        Yu Tian ; Kunbo Zhang ; Leyuan Wang ; Zhenan Sun
COMMENTS:       14pages,8figures
HIGHLIGHT:      In this paper, we present a face anti-spoofing method in a real-world scenario by automatic learning the physical characteristics in polarization images of a real face compared to a deceptive attack.

7, TITLE:       Global Texture Enhancement for Fake Face Detection in the Wild
http://arxiv.org/abs/2002.00133
AUTHORS:        Zhengzhe Liu ; Xiaojuan Qi ; Philip Torr
HIGHLIGHT:      In this paper, we conduct an empirical study on fake/real faces, and have two important observations: firstly, the texture of fake faces is substantially different from real ones; secondly, global texture statistics are more robust to image editing and transferable to fake faces from different GANs and datasets.

8, TITLE:       Batch-Shaping for Learning Conditional Channel Gated Networks
http://arxiv.org/abs/1907.06627
AUTHORS:        Babak Ehteshami Bejnordi ; Tijmen Blankevoort ; Max Welling
COMMENTS:       Published as a conference paper at ICLR 2020
HIGHLIGHT:      We present a method that trains large capacity neural networks with significantly improved accuracy and lower dynamic computational cost.

9, TITLE:       Detection of Pitt-Hopkins Syndrome based on morphological facial features
http://arxiv.org/abs/2003.08229
AUTHORS:        Elena D'Amato ; Constantino Carlos Reyes-Aldasoro ; Maria Felicia Faienza ; Marcella Zollino
COMMENTS:       Submitted to MIUA 2020
HIGHLIGHT:      This work describes an automatic methodology to discriminate between individuals with the genetic disorder Pitt-Hopkins syndrome (PTHS), and healthy individuals.

10, TITLE:      LT-Net: Label Transfer by Learning Reversible Voxel-wise Correspondence for One-shot Medical Image Segmentation
http://arxiv.org/abs/2003.07072
AUTHORS:        Shuxin Wang ; Shilei Cao ; Dong Wei ; Renzhen Wang ; Kai Ma ; Liansheng Wang ; Deyu Meng ; Yefeng Zheng
COMMENTS:       Accepted to Proc. IEEE Conf. Computer Vision and Pattern Recognition 2020
HIGHLIGHT:      We introduce a one-shot segmentation method to alleviate the burden of manual annotation for medical images.

11, TITLE:      Breast Cancer Detection Using Convolutional Neural Networks
http://arxiv.org/abs/2003.07911
AUTHORS:        Simon Hadush ; Yaecob Girmay ; Abiot Sinamo ; Gebrekirstos Hagos
HIGHLIGHT:      Deep learning techniques are revolutionizing the field of medical image analysis and hence in this study, we proposed Convolutional Neural Networks (CNNs) for breast mass detection so as to minimize the overheads of manual analysis.

12, TITLE:      PatchPerPix for Instance Segmentation
http://arxiv.org/abs/2001.07626
AUTHORS:        Peter Hirsch ; Lisa Mais ; Dagmar Kainmueller
HIGHLIGHT:      In this paper we present a novel method for proposal free instance segmentation that can handle sophisticated object shapes that span large parts of an image and form dense object clusters with crossovers.

13, TITLE:      CrypTFlow: Secure TensorFlow Inference
http://arxiv.org/abs/1909.07814
AUTHORS:        Nishant Kumar ; Mayank Rathee ; Nishanth Chandran ; Divya Gupta ; Aseem Rastogi ; Rahul Sharma
COMMENTS:       To appear at 41st IEEE Symposium on Security and Privacy 2020. Code available at: https://github.com/mpc-msri/EzPC
HIGHLIGHT:      We present CrypTFlow, a first of its kind system that converts TensorFlow inference code into Secure Multi-party Computation (MPC) protocols at the push of a button.

14, TITLE:      Datasheets for Datasets
http://arxiv.org/abs/1803.09010
AUTHORS:        Timnit Gebru ; Jamie Morgenstern ; Briana Vecchione ; Jennifer Wortman Vaughan ; Hanna Wallach ; Hal Daumé III ; Kate Crawford
COMMENTS:       Working Paper, comments are encouraged
HIGHLIGHT:      To address this gap, we propose datasheets for datasets.

15, TITLE:      PiiGAN: Generative Adversarial Networks for Pluralistic Image Inpainting
http://arxiv.org/abs/1912.01834
AUTHORS:        Weiwei Cai ; Zhanguo Wei
HIGHLIGHT:      In view of this weakness, which is related to the design of the previous algorithms, we propose a novel deep generative model equipped with a brand new style extractor which can extract the style feature (latent vector) from the ground truth.

16, TITLE:      BatVision: Learning to See 3D Spatial Layout with Two Ears
http://arxiv.org/abs/1912.07011
AUTHORS:        Jesper Haahr Christensen ; Sascha Hornauer ; Stella Yu
HIGHLIGHT:      We train a model to predict depth maps and even grayscale images from the sound alone.

17, TITLE:      Weakly and Semi Supervised Detection in Medical Imaging via Deep Dual Branch Net
http://arxiv.org/abs/1904.12589
AUTHORS:        Ran Bakalo ; Jacob Goldberger ; Rami Ben-Ari
HIGHLIGHT:      This study presents a novel deep learning architecture for multi-class classification and localization of abnormalities in medical imaging illustrated through experiments on mammograms.

18, TITLE:      Copy and Paste GAN: Face Hallucination from Shaded Thumbnails
http://arxiv.org/abs/2002.10650
AUTHORS:        Yang Zhang ; Ivor Tsang ; Yawei Luo ; Changhui Hu ; Xiaobo Lu ; Xin Yu
COMMENTS:       CVPR2020 (oral) preprint
HIGHLIGHT:      This paper proposes a Copy and Paste Generative Adversarial Network (CPGAN) to recover authentic high-resolution (HR) face images while compensating for low and non-uniform illumination.

19, TITLE:      Computing Maximum Matchings in Temporal Graphs
http://arxiv.org/abs/1905.05304
AUTHORS:        George B. Mertzios ; Hendrik Molter ; Rolf Niedermeier ; Viktor Zamaraev ; Philipp Zschoche
HIGHLIGHT:      We introduce and study the complexity of a natural temporal extension of the classical graph problem Maximum Matching, taking into account the dynamic nature of temporal graphs.

20, TITLE:      Limited Lookahead in Imperfect-Information Games
http://arxiv.org/abs/1902.06335
AUTHORS:        Christian Kroer ; Tuomas Sandholm
HIGHLIGHT:      We initiate a new direction via two simultaneous deviation points: generalization to imperfect-information games and a game-theoretic approach.

21, TITLE:      Mitigating Planner Overfitting in Model-Based Reinforcement Learning
http://arxiv.org/abs/1812.01129
AUTHORS:        Dilip Arumugam ; David Abel ; Kavosh Asadi ; Nakul Gopalan ; Christopher Grimm ; Jun Ki Lee ; Lucas Lehnert ; Michael L. Littman
HIGHLIGHT:      We present three different approaches that demonstrably mitigate planner overfitting in reinforcement-learning environments.

22, TITLE:      Fast Compressive Sensing Recovery Using Generative Models with Structured Latent Variables
http://arxiv.org/abs/1902.06913
AUTHORS:        Shaojie Xu ; Sihan Zeng ; Justin Romberg
HIGHLIGHT:      In this paper, we propose an algorithm for signal reconstruction from compressed measurements with image priors captured by a generative model.

23, TITLE:      Interpretable End-to-end Urban Autonomous Driving with Latent Deep Reinforcement Learning
http://arxiv.org/abs/2001.08726
AUTHORS:        Jianyu Chen ; Shengbo Eben Li ; Masayoshi Tomizuka
HIGHLIGHT:      In this paper, we propose an interpretable deep reinforcement learning method for end-to-end autonomous driving, which is able to handle complex urban scenarios.

24, TITLE:      Resolution Adaptive Networks for Efficient Inference
http://arxiv.org/abs/2003.07326
AUTHORS:        Le Yang ; Yizeng Han ; Xi Chen ; Shiji Song ; Jifeng Dai ; Gao Huang
COMMENTS:       CVPR 2020
HIGHLIGHT:      In this paper, we focus on spatial redundancy of input samples and propose a novel Resolution Adaptive Network (RANet), which is inspired by the intuition that low-resolution representations are sufficient for classifying "easy" inputs containing large objects with prototypical features, while only some "hard" samples need spatially detailed information.

25, TITLE:      Multi-Modal Domain Adaptation for Fine-Grained Action Recognition
http://arxiv.org/abs/2001.09691
AUTHORS:        Jonathan Munro ; Dima Damen
COMMENTS:       Accepted to CVPR 2020 for an oral presentation
HIGHLIGHT:      In this work we exploit the correspondence of modalities as a self-supervised alignment approach for UDA in addition to adversarial alignment.

26, TITLE:      Deep Neural Rejection against Adversarial Examples
http://arxiv.org/abs/1910.00470
AUTHORS:        Angelo Sotgiu ; Ambra Demontis ; Marco Melis ; Battista Biggio ; Giorgio Fumera ; Xiaoyi Feng ; Fabio Roli
HIGHLIGHT:      In this work, we propose a deep neural rejection mechanism to detect adversarial examples, based on the idea of rejecting samples that exhibit anomalous feature representations at different network layers.

27, TITLE:      Digitally Capturing Physical Prototypes During Early-Stage Engineering Design Projects for Initial Analysis of Project Output and Progression
http://arxiv.org/abs/1905.01950
AUTHORS:        Jorgen F. Erichsen ; Heikki Sjöman ; Martin Steinert ; Torgeir Welo
COMMENTS:       27 pages, 2 tables, 9 figures
HIGHLIGHT:      In this article, one project is shown in detail to demonstrate how this capturing system can gather empirical data for enriching engineering design project cases that focus on prototyping for concept generation.

28, TITLE:      Deep Parametric Shape Predictions using Distance Fields
http://arxiv.org/abs/1904.08921
AUTHORS:        Dmitriy Smirnov ; Matthew Fisher ; Vladimir G. Kim ; Richard Zhang ; Justin Solomon
COMMENTS:       Accepted to CVPR 2020
HIGHLIGHT:      Hence, we propose a new framework for predicting parametric shape primitives using deep learning.

29, TITLE:      Image Quality Transfer Enhances Contrast and Resolution of Low-Field Brain MRI in African Paediatric Epilepsy Patients
http://arxiv.org/abs/2003.07216
AUTHORS:        Matteo Figini ; Hongxiang Lin ; Godwin Ogbole ; Felice D Arco ; Stefano B. Blumberg ; David W. Carmichael ; Ryutaro Tanno ; Enrico Kaden ; Biobele J. Brown ; Ikeoluwa Lagunju ; Helen J. Cross ; Delmiro Fernandez-Reyes ; Daniel C. Alexander
COMMENTS:       6 pages, 3 figures, accepted at ICLR 2020 workshop on Artificial Intelligence for Affordable Healthcare
HIGHLIGHT:      Here we present qualitative results from real and simulated clinical low-field brain images showing the potential value of IQT to enhance the clinical utility of readily accessible low-field MRIs in the management of epilepsy.

30, TITLE:      Agriculture-Vision: A Large Aerial Image Database for Agricultural Pattern Analysis
http://arxiv.org/abs/2001.01306
AUTHORS:        Mang Tik Chiu ; Xingqian Xu ; Yunchao Wei ; Zilong Huang ; Alexander Schwing ; Robert Brunner ; Hrant Khachatrian ; Hovnatan Karapetyan ; Ivan Dozier ; Greg Rose ; David Wilson ; Adrian Tudor ; Naira Hovakimyan ; Thomas S. Huang ; Honghui Shi
COMMENTS:       CVPR 2020
HIGHLIGHT:      To encourage research in computer vision for agriculture, we present Agriculture-Vision: a large-scale aerial farmland image dataset for semantic segmentation of agricultural patterns.

31, TITLE:      Amora: Black-box Adversarial Morphing Attack
http://arxiv.org/abs/1912.03829
AUTHORS:        Run Wang ; Felix Juefei-Xu ; Xiaofei Xie ; Lei Ma ; Yihao Huang ; Yang Liu
HIGHLIGHT:      In this paper, we investigate and introduce a new type of adversarial attack to evade FR systems by manipulating facial content, called adversarial morphing attack (a.k.a. Amora).

32, TITLE:      MTI-Net: Multi-Scale Task Interaction Networks for Multi-Task Learning
http://arxiv.org/abs/2001.06902
AUTHORS:        Simon Vandenhende ; Stamatios Georgoulis ; Luc Van Gool
COMMENTS:       Updated version of the paper
HIGHLIGHT:      In this paper, we argue about the importance of considering task interactions at multiple scales when distilling task information in a multi-task learning setup.

33, TITLE:      Inexact Proximal-Point Penalty Methods for Constrained Non-Convex Optimization
http://arxiv.org/abs/1908.11518
AUTHORS:        Qihang Lin ; Runchao Ma ; Yangyang Xu
COMMENTS:       submitted to journal
HIGHLIGHT:      In this paper, an inexact proximal-point penalty method is studied for constrained optimization problems, where the objective function is non-convex, and the constraint functions can also be non-convex.

34, TITLE:      On the Importance of Word Order Information in Cross-lingual Sequence Labeling
http://arxiv.org/abs/2001.11164
AUTHORS:        Zihan Liu ; Genta Indra Winata ; Samuel Cahyawijaya ; Andrea Madotto ; Zhaojiang Lin ; Pascale Fung
COMMENTS:       11 pages, 1 figure, 7 tables
HIGHLIGHT:      In this paper, we hypothesize that cross-lingual models that fit into the word order of the source language might fail to handle target languages.

35, TITLE:      A Label Proportions Estimation Technique for Adversarial Domain Adaptation in Text Classification
http://arxiv.org/abs/2003.07444
AUTHORS:        Zhuohao Chen
COMMENTS:       correct typos, update the experiment results and add an appendix of proof
HIGHLIGHT:      In this study, we focus on unsupervised domain adaptation of text classification with label shift and introduce a domain adversarial network with label proportions estimation (DAN-LPE) framework.

36, TITLE:      Recent Advances and Challenges in Task-oriented Dialog System
http://arxiv.org/abs/2003.07490
AUTHORS:        Zheng Zhang ; Ryuichi Takanobu ; Minlie Huang ; Xiaoyan Zhu
COMMENTS:       Under review of SCIENCE CHINA Technological Science
HIGHLIGHT:      In this paper, we survey recent advances and challenges in an issue-specific manner.

37, TITLE:      CAiRE: An End-to-End Empathetic Chatbot
http://arxiv.org/abs/1907.12108
AUTHORS:        Zhaojiang Lin ; Peng Xu ; Genta Indra Winata ; Farhad Bin Siddique ; Zihan Liu ; Jamin Shin ; Pascale Fung
COMMENTS:       Extended version of AAAI-2020 demo paper
HIGHLIGHT:      In this paper, we present an end-to-end empathetic conversation agent CAiRE.

38, TITLE:      Sum of squares bounds for the ordering principle
http://arxiv.org/abs/1812.01163
AUTHORS:        Aaron Potechin
COMMENTS:       The details for the sum of squares lower bound have now been ironed out
HIGHLIGHT:      In this paper, we analyze the sum of squares hierarchy (SOS) on the ordering principle on $n$ elements.

39, TITLE:      ViewAL: Active Learning with Viewpoint Entropy for Semantic Segmentation
http://arxiv.org/abs/1911.11789
AUTHORS:        Yawar Siddiqui ; Julien Valentin ; Matthias Nießner
COMMENTS:       CVPR2020, Video: https://youtu.be/tAGdx2j-X_g
HIGHLIGHT:      We propose ViewAL, a novel active learning strategy for semantic segmentation that exploits viewpoint consistency in multi-view datasets.

40, TITLE:      Post-Training Piecewise Linear Quantization for Deep Neural Networks
http://arxiv.org/abs/2002.00104
AUTHORS:        Jun Fang ; Ali Shafiee ; Hamzah Abdel-Aziz ; David Thorsley ; Georgios Georgiadis ; Joseph Hassoun
HIGHLIGHT:      In this paper, we propose a piecewise linear quantization (PWLQ) scheme to enable accurate approximation for tensor values that have bell-shaped distributions with long tails.

41, TITLE:      Improved Embeddings with Easy Positive Triplet Mining
http://arxiv.org/abs/1904.04370
AUTHORS:        Hong Xuan ; Abby Stylianou ; Robert Pless
HIGHLIGHT:      In this paper, we propose an alternative, loosened embedding strategy that requires the embedding function only map each training image to the most similar examples from the same class, an approach we call "Easy Positive" mining.

42, TITLE:      Two Tier Prediction of Stroke Using Artificial Neural Networks and Support Vector Machines
http://arxiv.org/abs/2003.08354
AUTHORS:        Jerrin Thomas Panachakel ; Jeena R. S
COMMENTS:       arXiv admin note: text overlap with arXiv:1706.08227
HIGHLIGHT:      We have obtained an accuracy of 96.67% in tier-1 and an accuracy of 70% in tier-2.

43, TITLE:      Fairness in Deep Learning: A Computational Perspective
http://arxiv.org/abs/1908.08843
AUTHORS:        Mengnan Du ; Fan Yang ; Na Zou ; Xia Hu
HIGHLIGHT:      We provide a review covering recent progresses to tackle algorithmic fairness problems of deep learning from the computational perspective.

44, TITLE:      Neural Network Surgery with Sets
http://arxiv.org/abs/1912.06719
AUTHORS:        Jonathan Raiman ; Susan Zhang ; Christy Dennison
HIGHLIGHT:      We propose a solution to automatically determine which components of a neural network model should be salvaged and which require retraining.

45, TITLE:      Action Segmentation with Joint Self-Supervised Temporal Domain Adaptation
http://arxiv.org/abs/2003.02824
AUTHORS:        Min-Hung Chen ; Baopu Li ; Yingze Bao ; Ghassan AlRegib ; Zsolt Kira
COMMENTS:       CVPR 2020. Source code: https://github.com/cmhungsteve/SSTDA
HIGHLIGHT:      To reduce the discrepancy, we propose Self-Supervised Temporal Domain Adaptation (SSTDA), which contains two self-supervised auxiliary tasks (binary and sequential domain prediction) to jointly align cross-domain feature spaces embedded with local and global temporal dynamics, achieving better performance than other Domain Adaptation (DA) approaches.

46, TITLE:      Optimizing Medical Treatment for Sepsis in Intensive Care: from Reinforcement Learning to Pre-Trial Evaluation
http://arxiv.org/abs/2003.06474
AUTHORS:        Luchen Li ; Ignacio Albert-Smet ; Aldo A. Faisal
HIGHLIGHT:      Our aim is to establish a framework where reinforcement learning (RL) of optimizing interventions retrospectively allows us a regulatory compliant pathway to prospective clinical testing of the learned policies in a clinical deployment.

47, TITLE:      Defending Against Adversarial Examples with K-Nearest Neighbor
http://arxiv.org/abs/1906.09525
AUTHORS:        Chawin Sitawarin ; David Wagner
COMMENTS:       Inadequate experimental evaluation
HIGHLIGHT:      We propose a defense against adversarial examples based on a k-nearest neighbor (kNN) on the intermediate activation of neural networks.

48, TITLE:      FastSurfer -- A fast and accurate deep learning based neuroimaging pipeline
http://arxiv.org/abs/1910.03866
AUTHORS:        Leonie Henschel ; Sailesh Conjeti ; Santiago Estrada ; Kersten Diers ; Bruce Fischl ; Martin Reuter
COMMENTS:       Submitted to NeuroImage
HIGHLIGHT:      In this work we propose a fast and accurate deep learning based neuroimaging pipeline for the automated processing of structural human brain MRI scans, including surface reconstruction and cortical parcellation.

49, TITLE:      Weakly Supervised PET Tumor Detection Using Class Response
http://arxiv.org/abs/2003.08337
AUTHORS:        Amine Amyar ; Romain Modzelewski ; Pierre Vera ; Vincent Morard ; Su Ruan
COMMENTS:       Submitted to MICCAI 2020
HIGHLIGHT:      In this paper, we present a novel approach to locate different type of lesions in positron emission tomography (PET) images using only a class label at the image-level.

50, TITLE:      Learning for Video Compression with Hierarchical Quality and Recurrent Enhancement
http://arxiv.org/abs/2003.01966
AUTHORS:        Ren Yang ; Fabian Mentzer ; Luc Van Gool ; Radu Timofte
COMMENTS:       CVPR 2020 Camera-Ready
HIGHLIGHT:      In this paper, we propose a Hierarchical Learned Video Compression (HLVC) method with three hierarchical quality layers and a recurrent enhancement network.

51, TITLE:      Fully Unsupervised Probabilistic Noise2Void
http://arxiv.org/abs/1911.12291
AUTHORS:        Mangal Prakash ; Manan Lalit ; Pavel Tomancak ; Alexander Krull ; Florian Jug
COMMENTS:       Accepted at ISBI 2020
HIGHLIGHT:      Here, we present improvements to PN2V that (i) replace histogram based noise models by parametric noise models, and (ii) show how suitable noise models can be created even in the absence of calibration data.

52, TITLE:      Leveraging Self-supervised Denoising for Image Segmentation
http://arxiv.org/abs/1911.12239
AUTHORS:        Mangal Prakash ; Tim-Oliver Buchholz ; Manan Lalit ; Pavel Tomancak ; Florian Jug ; Alexander Krull
COMMENTS:       accepted at ISBI 2020
HIGHLIGHT:      Hence, we propose to use denoising networks to improve the performance of other DL-based image segmentation methods.

53, TITLE:      Conditional Convolutions for Instance Segmentation
http://arxiv.org/abs/2003.05664
AUTHORS:        Zhi Tian ; Chunhua Shen ; Hao Chen
HIGHLIGHT:      We propose a simple yet effective instance segmentation framework, termed CondInst (conditional convolutions for instance segmentation).

54, TITLE:      Dual-attention Guided Dropblock Module for Weakly Supervised Object Localization
http://arxiv.org/abs/2003.04719
AUTHORS:        Junhui Yin ; Siqing Zhang ; Dongliang Chang ; Zhanyu Ma ; Jun Guo
COMMENTS:       Technical Reports
HIGHLIGHT:      In this paper, we present a dual-attention guided dropblock module, and aim at learning the informative and complementary visual features for weakly supervised object localization (WSOL).