计算机视觉/图像处理学术速递[03.31]

格林先生MrGreen arXiv每日学术速递

cs.CV 方向,今日共计122篇


[检测分类相关]:
【1】 Quantifying the Scanner-Induced Domain Gap in Mitosis Detection
标题:有丝分裂检测中扫描仪诱导的区域间隙的量化
作者:Marc Aubreville,Christof Bertram,Mitko Veta,Robert Klopfleisch,Nikolas Stathonikos,Katharina Breininger,Natalie ter Hoeve,Francesco Ciompi,Andreas Maier
备注:3 pages, 1 figure, 1 table, submitted as short paper to MIDL
链接:arxiv.org/abs/2103.16515

【2】 Depth-conditioned Dynamic Message Propagation for Monocular 3D Object  Detection
标题:单目三维目标检测的深度约束动态信息传播算法
作者:Li Wang,Liang Du,Xiaoqing Ye,Yanwei Fu,Guodong Guo,Xiangyang Xue,Jianfeng Feng,Li Zhang
备注:CVPR 2021. Code at this https URL
链接:arxiv.org/abs/2103.16470

【3】 Beltrami Signature: A Novel Invariant 2D Shape Representation for Object  Classification
标题:Beltrami特征:一种新的用于目标分类的不变二维形状表示
作者:Chenran Lin,Lok Ming Lui
链接:arxiv.org/abs/2103.16411

【4】 Data-Uncertainty Guided Multi-Phase Learning for Semi-Supervised Object  Detection
标题:数据不确定性引导的多阶段学习半监督目标检测
作者:Zhenyu Wang,Yali Li,Ye Guo,Lu Fang,Shengjin Wang
备注:Accepted by CVPR 2021
链接:arxiv.org/abs/2103.16368

【5】 Delving into Localization Errors for Monocular 3D Object Detection
标题:单目三维目标检测中的定位误差研究
作者:Xinzhu Ma,Yinmin Zhang,Dan Xu,Dongzhan Zhou,Shuai Yi,Haojie Li,Wanli Ouyang
备注:CVPR'2021, code will be made available
链接:arxiv.org/abs/2103.16237

【6】 Class-Aware Robust Adversarial Training for Object Detection
标题:面向目标检测的类感知鲁棒对抗性训练
作者:Pin-Chun Chen,Bo-Han Kung,Jun-Cheng Chen
链接:arxiv.org/abs/2103.16148

【7】 Active Learning for Deep Object Detection via Probabilistic Modeling
标题:基于概率建模的主动学习在深度目标检测中的应用
作者:Jiwoong Choi,Ismail Elezi,Hyuk-Jae Lee,Clement Farabet,Jose M. Alvarez
链接:arxiv.org/abs/2103.16130

【8】 DeepWORD: A GCN-based Approach for Owner-Member Relationship Detection  in Autonomous Driving
标题:DeepWORD:一种基于GCN的自动驾驶车主关系检测方法
作者:Zizhang Wu,Man Wang,Jason Wang,Wenkai Zhang,Muqing Fang,Tianhao Xu
备注:Accepted by IEEE ICME
链接:arxiv.org/abs/2103.16099

【9】 3D-MAN: 3D Multi-frame Attention Network for Object Detection
标题:3D-MAN:用于目标检测的三维多帧注意力网络
作者:Zetong Yang,Yin Zhou,Zhifeng Chen,Jiquan Ngiam
链接:arxiv.org/abs/2103.16054

【10】 Revisiting Deep Local Descriptor for Improved Few-Shot Classification
标题:重新访问深度局部描述子进行改进的Few-Shot分类
作者:Jun He,Richang Hong,Xueliang Liu,Mingliang Xu,Meng Wang
备注:12 pages, 7 figures, 6 tables
链接:arxiv.org/abs/2103.16009

【11】 Detecting and Mapping Trees in Unstructured Environments with a Stereo  Camera and Pseudo-Lidar
标题:基于立体相机和伪激光雷达的非结构化环境中树木检测与映射
作者:Brian H. Wang,Carlos Diaz-Ruiz,Jacopo Banfi,Mark Campbell
备注:Accepted to the 2021 IEEE International Conference on Robotics and Automation (ICRA)
链接:arxiv.org/abs/2103.15967

[分割/语义相关]:
【1】 Boundary IoU: Improving Object-Centric Image Segmentation Evaluation
标题:边界借条:改进以对象为中心的图像分割评价
作者:Bowen Cheng,Ross Girshick,Piotr Dollár,Alexander C. Berg,Alexander Kirillov
备注:CVPR 2021, project page: this https URL
链接:arxiv.org/abs/2103.16562

【2】 Deep Gaussian Processes for Few-Shot Segmentation
标题:用于Few-Shot分割的深高斯过程
作者:Joakim Johnander,Johan Edstedt,Martin Danelljan,Michael Felsberg,Fahad Shahbaz Khan
备注:15 pages, 6 figures
链接:arxiv.org/abs/2103.16549

【3】 Source-Free Domain Adaptation for Semantic Segmentation
标题:一种无源域自适应的语义分词方法
作者:Yuang Liu,Wei Zhang,Jun Wang
备注:CVPR 2021, 10 pages
链接:arxiv.org/abs/2103.16372

【4】 Generalized Organ Segmentation by Imitating One-shot Reasoning using  Anatomical Correlation
标题:基于解剖相关的模仿一次推理的广义器官分割
作者:Hong-Yu Zhou,Hualuo Liu,Shilei Cao,Dong Wei,Chixiang Lu,Yizhou Yu,Kai Ma,Yefeng Zheng
备注:IPMI 2021
链接:arxiv.org/abs/2103.16344

【5】 Locate then Segment: A Strong Pipeline for Referring Image Segmentation
标题:先定位后分割:一条强有力的参考图像分割流水线
作者:Ya Jing,Tao Kong,Wei Wang,Liang Wang,Lei Li,Tieniu Tan
备注:CVPR 2021
链接:arxiv.org/abs/2103.16284

【6】 Multi-modal Trajectory Prediction for Autonomous Driving with Semantic  Map and Dynamic Graph Attention Network
标题:基于语义地图和动态图注意力网络的自动驾驶多模态轨迹预测
作者:Bo Dong,Hao Liu,Yu Bai,Jinbiao Lin,Zhuoran Xu,Xinyu Xu,Qi Kong
备注:NIPS2020 Workshop on Machine Learning for Autonomous Driving
链接:arxiv.org/abs/2103.16273

【7】 Is segmentation uncertainty useful?
标题:细分的不确定性有用吗?
作者:Steffen Czolbe,Kasra Arnavaz,Oswin Krause,Aasa Feragen
备注:Published at Information Processing in Medical Imaging (IPMI) 2021
链接:arxiv.org/abs/2103.16265

【8】 Multi-View Radar Semantic Segmentation
标题:多视点雷达语义分割
作者:Arthur Ouaknine,Alasdair Newson,Patrick Pérez,Florence Tupin,Julien Rebut
备注:15 pages, 8 figures. Preprint. Under review
链接:arxiv.org/abs/2103.16214

【9】 Self-Guided and Cross-Guided Learning for Few-Shot Segmentation
标题:用于Few-Shot分割的自引导学习和交叉引导学习
作者:Bingfeng Zhang,Jimin Xiao,Terry Qin
备注:CVPR 2021
链接:arxiv.org/abs/2103.16129

【10】 Assessing YOLACT++ for real time and robust instance segmentation of  medical instruments in endoscopic procedures
标题:评估YOLACT++在内窥镜手术中对医疗器械的实时和健壮的实例分割
作者:Juan Carlos Angeles Ceron,Leonardo Chang,Gilberto Ochoa-Ruiz,Sharib Ali
备注:Preprint under review for EMBC 2021 following IEEE guidelines
链接:arxiv.org/abs/2103.15997

【11】 DiNTS: Differentiable Neural Network Topology Search for 3D Medical  Image Segmentation
标题:DINTS:用于三维医学图像分割的可微神经网络拓扑搜索
作者:Yufan He,Dong Yang,Holger Roth,Can Zhao,Daguang Xu
备注:CVPR2021 oral
链接:arxiv.org/abs/2103.15954

【12】 Assessing the Role of Random Forests in Medical Image Segmentation
标题:随机森林在医学图像分割中的作用评估
作者:Dennis Hartmann,Dominik Müller,Iñaki Soto-Rey,Frank Kramer
链接:arxiv.org/abs/2103.16492

【13】 Automatic airway segmentation from Computed Tomography using robust and  efficient 3-D convolutional neural networks
标题:基于稳健高效三维卷积神经网络的CT气道自动分割
作者:A. Garcia-Uceda Juarez,R. Selvan,Z. Saghir,H. A. W. M. Tiddens,M. de Bruijne
链接:arxiv.org/abs/2103.16328

【14】 DualNorm-UNet: Incorporating Global and Local Statistics for Robust  Medical Image Segmentation
标题:DualNorm-UNET:融合全局和局部统计的鲁棒医学图像分割
作者:Junfei Xiao,Lequan Yu,Lei Xing,Alan Yuille,Yuyin Zhou
备注:code available at this https URL
链接:arxiv.org/abs/2103.15858

[人脸相关]:
【1】 Pre-training strategies and datasets for facial representation learning
标题:用于面部表征学习的预训练策略和数据集
作者:Adrian Bulat,Shiyang Cheng,Jing Yang,Andrew Garbett,Enrique Sanchez,Georgios Tzimiropoulos
链接:arxiv.org/abs/2103.16554

【2】 Face Forensics in the Wild
标题:面对野外取证
作者:Tianfei Zhou,Wenguan Wang,Zhiyuan Liang,Jianbing Shen
备注:CVPR 2021 (Oral). this https URL
链接:arxiv.org/abs/2103.16076

【3】 Identity-Aware CycleGAN for Face Photo-Sketch Synthesis and Recognition
标题:基于身份感知的人脸素描合成与识别CycleGan
作者:Yuke Fang,Jiani Hu,Weihong Deng
备注:36 pages, 11 figures
链接:arxiv.org/abs/2103.16019

【4】 High-fidelity Face Tracking for AR/VR via Deep Lighting Adaptation
标题:基于深光自适应的AR/VR高保真人脸跟踪
作者:Lele Chen,Chen Cao,Fernando De la Torre,Jason Saragih,Chenliang Xu,Yaser Sheikh
备注:The paper is accepted to CVPR 2021
链接:arxiv.org/abs/2103.15876

[GAN/对抗式/生成式相关]:
【1】 Enabling Data Diversity: Efficient Automatic Augmentation via  Regularized Adversarial Training
标题:实现数据多样性:通过正规化对抗性训练实现高效的自动增强
作者:Yunhe Gao,Zhiqiang Tang,Mu Zhou,Dimitris Metaxas
备注:Accepted by IPMI 2021
链接:arxiv.org/abs/2103.16493

【2】 What Causes Optical Flow Networks to be Vulnerable to Physical  Adversarial Attacks
标题:是什么导致光流网络易受物理攻击
作者:Simon Schrodi,Tonmoy Saikia,Thomas Brox
链接:arxiv.org/abs/2103.16255

【3】 SPatchGAN: A Statistical Feature Based Discriminator for Unsupervised  Image-to-Image Translation
标题:SPatchGAN:一种基于统计特征的无监督图像到图像翻译判别器
作者:Xuning Shao,Weidong Zhang
链接:arxiv.org/abs/2103.16219

【4】 Diagonal Attention and Style-based GAN for Content-Style Disentanglement  in Image Generation and Translation
标题:对角注意和基于风格的GAN在图像生成和翻译中的内容风格解缠
作者:Gihyun Kwon,Jong Chul Ye
链接:arxiv.org/abs/2103.16146

【5】 Automating Defense Against Adversarial Attacks: Discovery of  Vulnerabilities and Application of Multi-INT Imagery to Protect Deployed  Models
标题:自动防御敌方攻击:漏洞的发现和多点图像的应用保护已部署的模型
作者:Josh Kalin,David Noever,Matthew Ciolino,Dominick Hambrick,Gerry Dozier
备注:SPIE 2021, 8 Pages, 6 Figures
链接:arxiv.org/abs/2103.15897

【6】 Adversarially learned iterative reconstruction for imaging inverse  problems
标题:逆学习迭代重建成像反问题
作者:Subhadip Mukherjee,Ozan Öktem,Carola-Bibiane Schönlieb
备注:Accepted to the Eighth International Conference on Scale Space and Variational Methods in Computer Vision (SSVM), May-2021
链接:arxiv.org/abs/2103.16151

[图像/视频检索]:
【1】 Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with  Transformers
标题:思考快与慢:使用Transformers实现高效的文本到视觉检索
作者:Antoine Miech,Jean-Baptiste Alayrac,Ivan Laptev,Josef Sivic,Andrew Zisserman
备注:Accepted to CVPR 2021
链接:arxiv.org/abs/2103.16553

[行为/时空/光流/姿态/运动]:
【1】 Endo-Depth-and-Motion: Localization and Reconstruction in Endoscopic  Videos using Depth Networks and Photometric Constraints
标题:Endo-Depth-and-Motion:使用深度网络和光度约束在内窥镜视频中的定位和重建
作者:David Recasens,José Lamarca,José M. Fácil,J. M. M. Montiel,Javier Civera
链接:arxiv.org/abs/2103.16525

【2】 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment  Feedback Loop
标题:基于金字塔网格对齐反馈环的三维人体姿态和形状回归
作者:Hongwen Zhang,Yating Tian,Xinchi Zhou,Wanli Ouyang,Yebin Liu,Limin Wang,Zhenan Sun
备注:Technical report. Code and model available at this https URL
链接:arxiv.org/abs/2103.16507

【3】 Spatiotemporal Transformer for Video-based Person Re-identification
标题:基于视频的人的重新身份识别的时空变换
作者:Tianyu Zhang,Longhui Wei,Lingxi Xie,Zijie Zhuang,Yongfei Zhang,Bo Li,Qi Tian
备注:10 pages, 7 figures
链接:arxiv.org/abs/2103.16469

【4】 Graph Stacked Hourglass Networks for 3D Human Pose Estimation
标题:用于三维人体姿态估计的图叠式沙漏网络
作者:Tianhan Xu,Wataru Takano
备注:Accepted to CVPR 2021
链接:arxiv.org/abs/2103.16385

【5】 Learning monocular 3D reconstruction of articulated categories from  motion
标题:从运动中学习关节类别的单目三维重建
作者:Filippos Kokkinos,Iasonas Kokkinos
备注:For project website see this https URL
链接:arxiv.org/abs/2103.16352

【6】 Learning Parallel Dense Correspondence from Spatio-Temporal Descriptors  for Efficient and Robust 4D Reconstruction
标题:从时空描述符中学习并行稠密对应以实现高效和鲁棒的4D重建
作者:Jiapeng Tang,Dan Xu,Kui Jia,Lei Zhang
备注:15 pages, 11 figures, CVPR2021
链接:arxiv.org/abs/2103.16341

【7】 AGQA: A Benchmark for Compositional Spatio-Temporal Reasoning
标题:AGQA:一种构图时空推理基准
作者:Madeleine Grunde-McLaughlin,Ranjay Krishna,Maneesh Agrawala
备注:8 pages, 15 pages supplementary, 12 figures. To be published in CVPR 2021
链接:arxiv.org/abs/2103.16002

[半/弱/无监督相关]:
【1】 Broaden Your Views for Self-Supervised Video Learning
标题:开阔您的视野,了解自我指导的视频学习
作者:Adrià Recasens,Pauline Luc,Jean-Baptiste Alayrac,Luyu Wang,Florian Strub,Corentin Tallec,Mateusz Malinowski,Viorica Patraucean,Florent Altché,Michal Valko,Jean-Bastien Grill,Aäron van den Oord,Andrew Zisserman
链接:arxiv.org/abs/2103.16559

【2】 Unsupervised Learning of 3D Object Categories from Videos in the Wild
标题:野外视频中3D对象类别的无监督学习
作者:Philipp Henzler,Jeremy Reizenstein,Patrick Labatut,Roman Shapovalov,Tobias Ritschel,Andrea Vedaldi,David Novotny
链接:arxiv.org/abs/2103.16552

【3】 CoLA: Weakly-Supervised Temporal Action Localization with Snippet  Contrastive Learning
标题:COLA:基于片断对比学习的弱监督时间动作定位
作者:Can Zhang,Meng Cao,Dongming Yang,Jie Chen,Yuexian Zou
备注:Accepted by CVPR 2021
链接:arxiv.org/abs/2103.16392

【4】 ICE: Inter-instance Contrastive Encoding for Unsupervised Person  Re-identification
标题:ICE:无监督人员重识别的实例间对比编码
作者:Hao Chen,Benoit Lagadec,Francois Bremond
链接:arxiv.org/abs/2103.16364

【5】 MT3: Meta Test-Time Training for Self-Supervised Test-Time Adaption
标题:MT3:用于自监督测试时间自适应的元测试时间训练
作者:Alexander Bartler,Andre Bühler,Felix Wiewel,Mario Döbler,Bin Yang
链接:arxiv.org/abs/2103.16201

【6】 Weakly Supervised Temporal Action Localization Through Learning Explicit  Subspaces for Action and Context
标题:学习显式动作和上下文子空间的弱监督时间动作定位
作者:Ziyi Liu,Le Wang,Wei Tang,Junsong Yuan,Nanning Zheng,Gang Hua
备注:Accepted by the 35th AAAI Conference on Artificial Intelligence (AAAI 2021)
链接:arxiv.org/abs/2103.16155

【7】 Large Scale Autonomous Driving Scenarios Clustering with Self-supervised  Feature Extraction
标题:基于自监督特征提取的大规模自主驾驶场景聚类
作者:Jinxin Zhao,Jin Fang,Zhixian Ye,Liangjun Zhang
链接:arxiv.org/abs/2103.16101

【8】 Self-supervised Image-text Pre-training With Mixed Data In Chest X-rays
标题:胸片混合数据的自监督图文预训练
作者:Xiaosong Wang,Ziyue Xu,Leo Tam,Dong Yang,Daguang Xu
链接:arxiv.org/abs/2103.16022

【9】 Adaptive Pseudo-Label Refinement by Negative Ensemble Learning for  Source-Free Unsupervised Domain Adaptation
标题:基于负集成学习的无源无监督自适应伪标记域自适应
作者:Waqar Ahmed,Pietro Morerio,Vittorio Murino
链接:arxiv.org/abs/2103.15973

【10】 Tasting the cake: evaluating self-supervised generalization on  out-of-distribution multimodal MRI data
标题:尝一尝蛋糕:评估非分布多模态MRI数据的自我监督泛化
作者:Alex Fedorov,Eloy Geenjaar,Lei Wu,Thomas P. DeRamus,Vince D. Calhoun,Sergey M. Plis
备注:Accepted as a workshop paper at RobustML ICLR 2021
链接:arxiv.org/abs/2103.15914

[跟踪相关]:
【1】 Learning Target Candidate Association to Keep Track of What Not to Track
标题:学习目标候选人关联以跟踪不应跟踪的内容
作者:Christoph Mayer,Martin Danelljan,Danda Pani Paudel,Luc Van Gool
备注:17 Pages
链接:arxiv.org/abs/2103.16556

【2】 Learnable Graph Matching: Incorporating Graph Partitioning with Deep  Feature Learning for Multiple Object Tracking
标题:可学习图匹配:结合图分割和深度特征学习的多目标跟踪
作者:Jiawei He,Zehao Huang,Naiyan Wang,Zhaoxiang Zhang
备注:CVPR 2021 camera-ready version
链接:arxiv.org/abs/2103.16178

【3】 Dynamic Attention guided Multi-Trajectory Analysis for Single Object  Tracking
标题:动态注意力引导的单目标跟踪多轨迹分析
作者:Xiao Wang,Zhe Chen,Jin Tang,Bin Luo,Yaowei Wang,Yonghong Tian,Feng Wu
备注:Accepted by IEEE T-CSVT 2021
链接:arxiv.org/abs/2103.16086

[迁移学习/domain/主动学习/自适应]:
【1】 Bilevel Online Adaptation for Out-of-Domain Human Mesh Reconstruction
标题:基于双层在线自适应的域外人体网格重建
作者:Shanyan Guan,Jingwei Xu,Yunbo Wang,Bingbing Ni,Xiaokang Yang
备注:CVPR 2021, the project page: this https URL
链接:arxiv.org/abs/2103.16449

【2】 Dynamic Domain Adaptation for Efficient Inference
标题:用于高效推理的动态域自适应
作者:Shuang Li,Jinming Zhang,Wenxuan Ma,Chi Harold Liu,Wei Li
备注:Accepted by CVPR 2021
链接:arxiv.org/abs/2103.16403

【3】 Leveraging Self-Supervision for Cross-Domain Crowd Counting
标题:利用自我监督进行跨域人群清点
作者:Weizhe Liu,Nikita Durasov,Pascal Fua
链接:arxiv.org/abs/2103.16291

【4】 Two-Stage Monte Carlo Denoising with Adaptive Sampling and Kernel Pool
标题:基于自适应采样和核池的两阶段蒙特卡罗去噪
作者:Tiange Xiang,Hongliang Yuan,Haozhi Huang,Yujin Shi
链接:arxiv.org/abs/2103.16115

【5】 Kaleido-BERT: Vision-Language Pre-training on Fashion Domain
标题:Kaleido-Bert:时尚领域的视觉语言预训
作者:Mingchen Zhuge,Dehong Gao,Deng-Ping Fan,Linbo Jin,Ben Chen,Haoming Zhou,Minghui Qiu,Ling Shao
备注:CVPR2021 Accepted. Code: this https URL
链接:arxiv.org/abs/2103.16110

【6】 Progressive Domain Expansion Network for Single Domain Generalization
标题:用于单域泛化的渐进域扩展网络
作者:Lei Li,Ke Gao,Juan Cao,Ziyao Huang,Yepeng Weng,Xiaoyue Mi,Zhengze Yu,Xiaoya Li,Boyang xia
备注:Accepted to CVPR2021
链接:arxiv.org/abs/2103.16050

【7】 Augmented Transformer with Adaptive Graph for Temporal Action Proposal  Generation
标题:基于自适应图的增广Transformer时间行动方案生成
作者:Shuning Chang,Pichao Wang,Fan Wang,Hao Li,Jiashi Feng
备注:12 pagess, 4 figures
链接:arxiv.org/abs/2103.16024

【8】 Domain-robust VQA with diverse datasets and methods but no target labels
标题:具有多种数据集和方法但没有目标标签的域健壮VQA
作者:Mingda Zhang,Tristan Maidment,Ahmad Diab,Adriana Kovashka,Rebecca Hwa
备注:To appear in CVPR 2021
链接:arxiv.org/abs/2103.15974

【9】 Learning Domain Invariant Representations for Generalizable Person  Re-Identification
标题:基于学习域不变表示的泛化人物再识别
作者:Yi-Fan Zhang,Hanlin Zhang,Zhang Zhang,Da Li,Zhen Jia,Liang Wang,Tieniu Tan
链接:arxiv.org/abs/2103.15890

[数据集dataset]:
【1】 Automated Cleanup of the ImageNet Dataset by Model Consensus,  Explainability and Confident Learning
标题:基于模型一致性、可解释性和可信度学习的ImageNet数据集自动清理
作者:Csaba Kertész
链接:arxiv.org/abs/2103.16324

【2】 Does it work outside this benchmark? Introducing the Rigid Depth  Constructor tool, depth validation dataset construction in rigid scenes for  the masses
标题:它在这个基准之外工作吗?引入刚性深度构造器工具,为群众构建刚性场景中的深度验证数据集
作者:Clément Pinard,Antoine Manzanera
链接:arxiv.org/abs/2103.15970

[超分辨率]:
【1】 Flow-based Kernel Prior with Application to Blind Super-Resolution
标题:基于流的核先验及其在盲超分辨中的应用
作者:Jingyun Liang,Kai Zhang,Shuhang Gu,Luc Van Gool,Radu Timofte
备注:Accepted by CVPR2021. Code: this https URL
链接:arxiv.org/abs/2103.15977

[点云]:
【1】 Free-form Description Guided 3D Visual Graph Network for Object  Grounding in Point Cloud
标题:自由形式描述制导的三维视觉图形网络在点云对象固定中的应用
作者:Mingtao Feng,Zhen Li,Qi Li,Liang Zhang,XiangDong Zhang,Guangming Zhu,Hui Zhang,Yaonan Wang,Ajmal Mian
链接:arxiv.org/abs/2103.16381

【2】 PointBA: Towards Backdoor Attacks in 3D Point Cloud
标题:PointBA:针对三维点云中的后门攻击
作者:Xinke Li,Zhiru Chen,Yue Zhao,Zekun Tong,Yabang Zhao,Andrew Lim,Joey Tianyi Zhou
链接:arxiv.org/abs/2103.16074

【3】 Fast and Accurate Normal Estimation for Point Cloud via Patch Stitching
标题:基于面片拼接的快速准确的点云法线估计
作者:Jun Zhou,Wei Jin,Mingjie Wang,Xiuping Liu,Zhiyang Li,Zhaobin Liu
链接:arxiv.org/abs/2103.16066

[深度depth相关]:
【1】 Physics-based Differentiable Depth Sensor Simulation
标题:基于物理的差动深度传感器仿真
作者:Benjamin Planche,Rajat Vikram Singh
链接:arxiv.org/abs/2103.16563

[3D/3D重建等相关]:
【1】 3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding
标题:3D AffordanceNet:视觉对象听觉理解的基准
作者:Shengheng Deng,Xun Xu,Chaozheng Wu,Ke Chen,Kui Jia
链接:arxiv.org/abs/2103.16397

【2】 Deep regression on manifolds: a 3D rotation case study
标题:流形上的深度回归:三维旋转案例研究
作者:Romain Brégier
链接:arxiv.org/abs/2103.16317

【3】 Reconstructing Interactive 3D Scenes by Panoptic Mapping and CAD Model  Alignments
标题:基于全景映射和CAD模型对齐的交互式三维场景重建
作者:Muzhi Han,Zeyu Zhang,Ziyuan Jiao,Xu Xie,Yixin Zhu,Song-Chun Zhu,Hangxin Liu
备注:ICRA 2021 paper. Project: this https URL
链接:arxiv.org/abs/2103.16095

[OCR]:
【1】 A Multiplexed Network for End-to-End, Multilingual OCR
标题:一种端到端、多语言OCR的多路复用网络
作者:Jing Huang,Guan Pang,Rama Kovvuri,Mandy Toh,Kevin J Liang,Praveen Krishnan,Xi Yin,Tal Hassner
链接:arxiv.org/abs/2103.15992

[其他视频相关]:
【1】 Recognizing Actions in Videos from Unseen Viewpoints
标题:从看不见的角度识别视频中的动作
作者:AJ Piergiovanni,Michael S. Ryoo
备注:None
链接:arxiv.org/abs/2103.16516

【2】 Read and Attend: Temporal Localisation in Sign Language Videos
标题:阅读和参与:手语视频中的时间本地化
作者:Gül Varol,Liliane Momeni,Samuel Albanie,Triantafyllos Afouras,Andrew Zisserman
备注:Appears in: 2021 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2021). 14 pages
链接:arxiv.org/abs/2103.16481

【3】 Temporal Memory Relation Network for Workflow Recognition from Surgical  Video
标题:时态记忆关系网络在手术视频工作流识别中的应用
作者:Yueming Jin,Yonghao Long,Cheng Chen,Zixu Zhao,Qi Dou,Pheng-Ann Heng
备注:Accepted at IEEE Transactions on Medical Imaging (IEEE TMI); Code is available at this https URL
链接:arxiv.org/abs/2103.16327

【4】 Head2HeadFS: Video-based Head Reenactment with Few-shot Learning
标题:Head2HeadFS:基于视频的Few-Shot学习的头部重现
作者:Michail Christos Doukas,Mohammad Rami Koujan,Viktoriia Sharmanska,Stefanos Zafeiriou
链接:arxiv.org/abs/2103.16229

【5】 XVFI: eXtreme Video Frame Interpolation
标题:XVFI:极限视频帧插值
作者:Hyeonjun Sim,Jihyong Oh,Munchurl Kim
备注:The first two authors contributed equally to this work
链接:arxiv.org/abs/2103.16206

[其他]:
【1】 Learning Representational Invariances for Data-Efficient Action  Recognition
标题:面向数据高效动作识别的表征不变性学习
作者:Yuliang Zou,Jinwoo Choi,Qitong Wang,Jia-Bin Huang
备注:Project page: this https URL
链接:arxiv.org/abs/2103.16565

【2】 Grounding Physical Concepts of Objects and Events Through Dynamic Visual  Reasoning
标题:通过动态视觉推理使对象和事件的物理概念落地
作者:Zhenfang Chen,Jiayuan Mao,Jiajun Wu,Kwan-Yee Kenneth Wong,Joshua B. Tenenbaum,Chuang Gan
备注:ICLR 2021. Project page: this http URL
链接:arxiv.org/abs/2103.16564

【3】 Diagnosing Vision-and-Language Navigation: What Really Matters
标题:诊断视觉和语言导航:什么才是真正重要的
作者:Wanrong Zhu,Yuankai Qi,Pradyumna Narayana,Kazoo Sone,Sugato Basu,Xin Eric Wang,Qi Wu,Miguel Eckstein,William Yang Wang
链接:arxiv.org/abs/2103.16561

【4】 The Elastic Lottery Ticket Hypothesis
标题:弹性彩票假说
作者:Xiaohan Chen,Yu Cheng,Shuohang Wang,Zhe Gan,Jingjing Liu,Zhangyang Wang
链接:arxiv.org/abs/2103.16547

【5】 Visual Room Rearrangement
标题:视觉房间重新排列
作者:Luca Weihs,Matt Deitke,Aniruddha Kembhavi,Roozbeh Mottaghi
备注:CVPR 2021 - Oral Presentation
链接:arxiv.org/abs/2103.16544

【6】 SD-6DoF-ICLK: Sparse and Deep Inverse Compositional Lucas-Kanade  Algorithm on SE(3)
标题:SD-6DoF-ICLK:SE(3)上的稀疏深逆组合Lucas-Kanade算法
作者:Timo Hinzmann,Roland Siegwart
备注:Initial submission; 7 pages, 3 figures, 2 tables
链接:arxiv.org/abs/2103.16528

【7】 HapTable: An Interactive Tabletop Providing Online Haptic Feedback for  Touch Gestures
标题:HapTable:一个为触摸手势提供在线触觉反馈的交互式桌面
作者:Senem Ezgi Emgin,Amirreza Aghakhani,T. Metin Sezgin,Cagatay Basdogan
备注:None
链接:arxiv.org/abs/2103.16510

【8】 Benchmarking Representation Learning for Natural World Image Collections
标题:自然世界图像集的基准表征学习
作者:Grant Van Horn,Elijah Cole,Sara Beery,Kimberly Wilber,Serge Belongie,Oisin Mac Aodha
备注:CVPR 2021
链接:arxiv.org/abs/2103.16483

【9】 SIMstack: A Generative Shape and Instance Model for Unordered Object  Stacks
标题:SIMstack:一种面向无序对象堆栈的生成式形状和实例模型
作者:Zoe Landgraf,Raluca Scona,Tristan Laidlow,Stephen James,Stefan Leutenegger,Andrew J. Davison
链接:arxiv.org/abs/2103.16442

【10】 Causal Hidden Markov Model for Time Series Disease Forecasting
标题:时间序列疾病预测的因果隐马尔可夫模型
作者:Jing Li,Botong Wu,Xinwei Sun,Yizhou Wang
链接:arxiv.org/abs/2103.16391

【11】 Distribution Alignment: A Unified Framework for Long-tail Visual  Recognition
标题:分布对齐:一种长尾视觉识别的统一框架
作者:Songyang Zhang,Zeming Li,Shipeng Yan,Xuming He,Jian Sun
备注:Accepted by CVPR 2021
链接:arxiv.org/abs/2103.16370

【12】 Complementary Relation Contrastive Distillation
标题:互补关系对比蒸馏
作者:Jinguo Zhu,Shixiang Tang,Dapeng Chen,Shijie Yu,Yakun Liu,Aijun Yang,Mingzhe Rong,Xiaohua Wang
备注:CVPR2021 Poster
链接:arxiv.org/abs/2103.16367

【13】 Foveated Neural Radiance Fields for Real-Time and Egocentric Virtual  Reality
标题:用于实时、以自我为中心的虚拟现实的凹陷神经辐射场
作者:Nianchen Deng,Zhenyi He,Jiannan Ye,Praneeth Chakravarthula,Xubo Yang,Qi Sun
链接:arxiv.org/abs/2103.16365

【14】 Differentiable Network Adaption with Elastic Search Space
标题:弹性搜索空间下的可微网络自适应
作者:Shaopeng Guo,Yujie Wang,Kun Yuan,Quanquan Li
链接:arxiv.org/abs/2103.16350

【15】 Rethinking Spatial Dimensions of Vision Transformers
标题:对视觉Transformer空间维度的再思考
作者:Byeongho Heo,Sangdoo Yun,Dongyoon Han,Sanghyuk Chun,Junsuk Choe,Seong Joon Oh
备注:10 pages, 5 figures
链接:arxiv.org/abs/2103.16302

【16】 Single Test Image-Based Automated Machine Learning System for  Distinguishing between Trait and Diseased Blood Samples
标题:基于单次检测图像的血样性状与病态识别自动机器学习系统
作者:Sahar A. Nasser,Debjani Paul,Suyash P. Awate
链接:arxiv.org/abs/2103.16285

【17】 Model-Contrastive Federated Learning
标题:模型对比联邦学习
作者:Qinbin Li,Bingsheng He,Dawn Song
备注:Accepted by CVPR 2021
链接:arxiv.org/abs/2103.16257

【18】 Improving robustness against common corruptions with frequency biased  models
标题:利用频率偏差模型提高对常见腐败的鲁棒性
作者:Tonmoy Saikia,Cordelia Schmid,Thomas Brox
链接:arxiv.org/abs/2103.16241

【19】 Using Low-rank Representation of Abundance Maps and Nonnegative Tensor  Factorization for Hyperspectral Nonlinear Unmixing
标题:利用丰度图的低秩表示和非负张量分解实现高光谱非线性分解
作者:Lianru Gao,Zhicheng Wang,Lina Zhuang,Haoyang Yu,Bing Zhang,Jocelyn Chanussot
链接:arxiv.org/abs/2103.16204

【20】 Differentiable Drawing and Sketching
标题:微分绘画与素描
作者:Daniela Mihai,Jonathon Hare
链接:arxiv.org/abs/2103.16194

【21】 Repopulating Street Scenes
标题:重新填充街道场景
作者:Yifan Wang,Andrew Liu,Richard Tucker,Jiajun Wu,Brian L. Curless,Steven M. Seitz,Noah Snavely
备注:CVPR 2021
链接:arxiv.org/abs/2103.16183

【22】 Contrastive Embedding for Generalized Zero-Shot Learning
标题:广义零点学习的对比嵌入算法
作者:Zongyan Han,Zhenyong Fu,Shuo Chen,Jian Yang
备注:Accepted by CVPR2021
链接:arxiv.org/abs/2103.16173

【23】 FONTNET: On-Device Font Understanding and Prediction Pipeline
标题:FONTNET:设备内字体理解和预测管道
作者:Rakshith S,Rishabh Khurana,Vibhav Agarwal,Jayesh Rajkumar Vachhani,Guggilla Bhanodai
备注:Accepted for publication in IEEE ICASSP 2021: 46th IEEE International Conference on Acoustics, Speech, & Signal Processing
链接:arxiv.org/abs/2103.16150

【24】 Large Scale Visual Food Recognition
标题:大规模视觉食品识别
作者:Weiqing Min,Zhiling Wang,Yuxin Liu,Mengjiang Luo,Liping Kang,Xiaoming Wei,Xiaolin Wei,Shuqiang Jiang
链接:arxiv.org/abs/2103.16107

【25】 Deep Learning and Machine Vision for Food Processing: A Survey
标题:深度学习与机器视觉在食品加工中的研究进展
作者:Lili Zhu,Petros Spachos,Erica Pensini,Konstantinos Plataniotis
链接:arxiv.org/abs/2103.16106

【26】 Fully Convolutional Scene Graph Generation
标题:全卷积场景图的生成
作者:Hengyue Liu,Ning Yan,Masood S. Mortazavi,Bir Bhanu
备注:CVPR 2021 Oral
链接:arxiv.org/abs/2103.16083

【27】 Environmental sound analysis with mixup based multitask learning and  cross-task fusion
标题:基于混合多任务学习和跨任务融合的环境声分析
作者:Weiping Zheng,Dacan Jiang,Gansen Zhao
备注:5 pages, 1 figue
链接:arxiv.org/abs/2103.16079

【28】 Noise-resistant Deep Metric Learning with Ranking-based Instance  Selection
标题:基于排序的实例选择抗噪深度度量学习
作者:Chang Liu,Han Yu,Boyang Li,Zhiqi Shen,Zhanning Gao,Peiran Ren,Xuansong Xie,Lizhen Cui,Chunyan Miao
备注:Accepted by CVPR 2021
链接:arxiv.org/abs/2103.16047

【29】 Progressively Complementary Network for Fisheye Image Rectification  Using Appearance Flow
标题:基于外观流的渐进式互补网络鱼眼图像纠正
作者:Shangrong Yang,Chunyu Lin,Kang Liao,Chunjie Zhang,Yao Zhao
备注:10 pages, 12 figures
链接:arxiv.org/abs/2103.16026

【30】 Training Sparse Neural Network by Constraining Synaptic Weight on Unit  Lp Sphere
标题:单位Lp球面上约束突触权重训练稀疏神经网络
作者:Weipeng Li,Xiaogang Yang,Chuanxiang Li,Ruitao Lu,Xueli Xie
链接:arxiv.org/abs/2103.16013

【31】 TransFill: Reference-guided Image Inpainting by Merging Multiple Color  and Spatial Transformations
标题:TransFill:融合多种颜色和空间变换的参考引导图像修复
作者:Yuqian Zhou,Connelly Barnes,Eli Shechtman,Sohrab Amirghodsi
备注:Accepted by CVPR2021
链接:arxiv.org/abs/2103.15982

【32】 A tutorial on $\mathbf{SE}(3)$ transformation parameterizations and  on-manifold optimization
作者:José Luis Blanco-Claraco
备注:68 pages, 6 figures; v1 in arXiv; see history of document versions on page 3 for full change log of the technical report since 2010
链接:arxiv.org/abs/2103.15980

【33】 A Simple Approach for Zero-Shot Learning based on Triplet Distribution  Embeddings
标题:一种基于三重态分布嵌入的简单零射学习方法
作者:Vivek Chalumuri,Bac Nguyen
链接:arxiv.org/abs/2103.15939

【34】 Online Defense of Trojaned Models using Misattributions
标题:使用错误属性的特洛伊木马模型的在线防御
作者:Panagiota Kiourti,Wenchao Li,Anirban Roy,Karan Sikka,Susmit Jha
链接:arxiv.org/abs/2103.15918

【35】 Robust Audio-Visual Instance Discrimination
标题:稳健的视听实例识别
作者:Pedro Morgado,Ishan Misra,Nuno Vasconcelos
链接:arxiv.org/abs/2103.15916

【36】 Sign Language Production: A Review
标题:手语产生研究述评
作者:Razieh Rastgoo,Kourosh Kiani,Sergio Escalera,Mohammad Sabokrou
链接:arxiv.org/abs/2103.15910

【37】 Comparison of different convolutional neural network activa-tion  functions and methods for building ensembles
标题:构建集成的不同卷积神经网络激活函数和方法的比较
作者:Loris Nanni,Gianluca Maguolo,Sheryl Brahnam,Michelangelo Paci
链接:arxiv.org/abs/2103.15898

【38】 In-Place Scene Labelling and Understanding with Implicit Scene  Representation
标题:基于隐式场景表示的就地场景标注与理解
作者:Shuaifeng Zhi,Tristan Laidlow,Stefan Leutenegger,Andrew J. Davison
备注:Project page with more videos: this https URL
链接:arxiv.org/abs/2103.15875

【39】 Is Image-to-Image Translation the Panacea for Multimodal Image  Registration? A Comparative Study
标题:图像到图像翻译是多模式图像配准的灵丹妙药吗?比较研究
作者:Jiahao Lu,Johan Öfverstedt,Joakim Lindblad,Nataša Sladoje
备注:32 pages, 7 figures
链接:arxiv.org/abs/2103.16262

【40】 Machine learning method for light field refocusing
标题:一种用于光场重聚焦的机器学习方法
作者:Eisa Hedayati,Timothy C. Havens,Jeremy P. Bos
链接:arxiv.org/abs/2103.16020

【41】 Iterative Gradient Encoding Network with Feature Co-Occurrence Loss for  Single Image Reflection Removal
标题:用于单幅图像反射去除的特征共生损失迭代梯度编码网络
作者:Sutanu Bera,Prabir Kumar Biswas
备注:Submitted to IEEE International Conference of Image Processing (ICIP)
链接:arxiv.org/abs/2103.15903

机器翻译,仅供参考
点击阅读原文访问www.arxivdaily.com,获取带摘要及更多学科的学术速递。

文章推荐