视觉论文以及开源库
CVPR2019 全部论文汇总：
https://github.com/extreme-assistant/cvpr2019
CVPR2019 论文解读
http://bbs.cvmart.net/topics/287/cvpr2019
1.Graph Attention Convolution for Point Cloud Segmentation
论文链接：https://engineering.purdue.edu/~jshan/publications/2018/Lei Wang Graph Attention Convolution for Point Cloud Segmentation CVPR2019.pdf
2.GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in Point Cloud
作者：Li Yi, Wang Zhao, He Wang, Minhyuk Sung, Leonidas Guibas
论文链接：https://arxiv.org/abs/1812.03320
3.Nesti-Net: Normal Estimation for Unstructured 3D Point Clouds using Convolutional Neural Networks
作者：Yizhak Ben-Shabat, Michael Lindenbaum, Anath Fischer
论文链接：https://arxiv.org/abs/1812.00709
源码链接：https://github.com/sitzikbs/Nesti-Net
4.Robust Point Cloud Based Reconstruction of Large-Scale Outdoor Scenes
作者：Ziquan Lan, Zi Jian Yew，Gim Hee Lee
论文链接：https://www.researchgate.net/publication/332240602_Robust_Point_Cloud_Based_Reconstruction_of_Large-Scale_Outdoor_Scenes
源码链接：https://github.com/ziquan111/RobustPCLReconstruction
5.PointNetLK: Robust & Efficient Point Cloud Registration using PointNet
作者：Yasuhiro Aoki, Hunter Goforth, Rangaprasad Arun Srivatsan, Simon Lucey
论文链接：https://arxiv.org/abs/1903.05711
源码链接：https://github.com/hmgoforth/PointNetLK
6.PointWeb: Enhancing Local Neighborhood Features for Point Cloud Processing
作者：Hengshuang Zhao， Li Jiang， Chi-Wing Fu，Jiaya Jia
论文链接：http://jiaya.me/papers/pointweb_cvpr19.pdf
7.ClusterNet: Deep Hierarchical Cluster Network with Rigorously Rotation-Invariant Representation for Point Cloud Analysis
作者：Chao Chen， Guanbin Li， Ruijia Xu， Tianshui Chen， Meng Wang， Liang Lin
论文链接：http://www.linliang.net/wp-content/uploads/2019/04/CVPR2019_PointClound.pdf
8.FilterReg: Robust and Efficient Probabilistic Point-Set Registration using Gaussian Filter and Twist Parameterization
作者：Wei Gao, Russ Tedrake
论文链接：https://arxiv.org/abs/1811.10136
源码链接：https://bitbucket.org/gaowei19951004/poser/src/master/
9.Embodied Question Answering in Photorealistic Environments with Point Cloud Perception
作者：Erik Wijmans, Samyak Datta, Oleksandr Maksymets, Abhishek Das, Georgia Gkioxari, Stefan Lee, Irfan Essa, Devi Parikh, Dhruv Batra
论文链接：https://arxiv.org/abs/1904.03461
10.SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondences
作者：Huu Le, Thanh-Toan Do, Tuan Hoang, Ngai-Man Cheung
论文链接：https://arxiv.org/abs/1904.03483
源码链接：https://github.com/intellhave/SDRSAC （matlab）
11.PointFlowNet: Learning Representations for Rigid Motion Estimation from Point Clouds
作者：Aseem Behl, Despoina Paschalidou, Simon Donné, Andreas Geiger
论文链接：https://arxiv.org/abs/1806.02170
源码链接：https://github.com/aseembehl/pointflownet
12.PointPillars: Fast Encoders for Object Detection from Point Clouds
作者：Alex H. Lang, Sourabh Vora, Holger Caesar, Lubing Zhou, Jiong Yang, Oscar Beijbom
论文链接：https://arxiv.org/abs/1812.05784
源码链接：https://github.com/nutonomy/second.pytorch
13.PointNetLK: Point Cloud Registration using PointNet
作者：Yasuhiro Aoki, Hunter Goforth, Rangaprasad Arun Srivatsan, Simon Lucey
论文链接：https://arxiv.org/abs/1903.05711
源码链接：https://github.com/hmgoforth/PointNetLK
14.Supervised Fitting of Geometric Primitives to 3D Point Clouds（Oral）
作者：Lingxiao Li, Minhyuk Sung, Anastasia Dubrovina, Li Yi, Leonidas Guibas
论文链接：https://arxiv.org/abs/1811.08988
源码链接：https://github.com/csimstu2/SPFN
15.PointConv: Deep Convolutional Networks on 3D Point Clouds
作者：Wenxuan Wu, Zhongang Qi, Li Fuxin
论文链接：https://arxiv.org/abs/1811.07246
源码链接：https://github.com/DylanWusee/pointconv
16.Modeling Point Clouds with Self-Attention and Gumbel Subset Sampling
作者：Jiancheng Yang, Qiang Zhang, Bingbing Ni, Linguo Li, Jinxian Liu, Mengdie Zhou, Qi Tian
论文链接：https://arxiv.org/abs/1904.03375v1
17.Spherical Fractal Convolutional Neural Networks for Point Cloud Recognition
作者：Yongming Rao, Jiwen Lu, Jie Zhou
论文链接：https://raoyongming.github.io/files/SFCNN.pdf
OpenVSLAM：日本先进工业科技研究所新开源视觉SLAM框架
5月20日，日本先进工业科技研究所（National Institute of Advanced Industrial Science and Technology ）开源了一套视觉SLAM算法：OpenVSLAM。
开源地址：
https://github.com/xdspacelab/openvslam
OpenVSLAM是一套单目、立体、RGB-D视觉SLAM系统，其主要特点：
兼容多种相机类型，并可以轻松定制兼容其他类型相机；
可以存储和加载创建的地图，然后OpenVSLAM可以基于预先构建的地图定位新图像；
系统完全模块化的；
提供了一些代码片段来理解该系统的核心功能。
官方提供了较详细的文档：
https://openvslam.readthedocs.io/en/master/
谷歌大脑提出EfficientNet平衡模型扩展三个维度，取得精度-效率的最大化！
谷歌大脑新出的论文《EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks》，在模型扩展时平衡好深度、宽度、分辨率，取得精度、效率、模型大小的最大化。借由此简单有效的模型扩展方法，作者在使用神经架构搜索得到的基模型上扩展出一系列模型EfficientNets，达到了更好的精度和效率的平衡，其中EfficientNet-B7模型在ImageNet数据集上达到 state-of-the-art 84.4% top-1 / 97.1% top-5 精度，并且相比目前最好的方法模型size减小8.4倍，速度快6.1！！简直是神级操作！该文已被ICML 2019录用，这可能是一篇要改变整个深度卷积网络模型设计的论文。
论文地址：
https://arxiv.org/pdf/1905.11946v1.pdf
开源地址：
https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet
计算机视觉开源论文
MobileNetV3 的开源犹如一阵旋风，突然出现了几十个项目。
C3F 的开源使得人群计数领域终于有了自己的框架。
还有语义分割、目标跟踪、表情识别、姿态估计、超分辨率等的开源代码，他们来自最近新出的论文，对于相关方向的同学肯定很有借鉴意义。
谷歌最新MobileNetV3开源实现！
https://github.com/xiaolai-sqlai/mobilenetv3
C3F：首个开源人群计数算法框架
https://github.com/gjy3035/C-3-Framework
快到没朋友的YOLOv3有了PaddlePaddle 预训练模型，精度更高了！
https://github.com/PaddlePaddle/models/blob/v1.4/PaddleCV/yolov3/README_cn.md
语义分割 | 高效的梯形风格的DenseNets网络LDN，用于大图像的语义分割
Efficient Ladder-style DenseNets for Semantic Segmentation of Large Images
Ivan Krešo, Josip Krapac, Siniša Šegvić
https://arxiv.org/abs/1905.05661v1
（代码将于论文被接收后开源，还未公布地址）
目标跟踪 | 研究从相关滤波跟踪算法中去除常见的Cosine Window机制,与传统算法和深度学习算法相比都取得了不错的精度。
Remove Cosine Window from Correlation Filter-based Visual Trackers: When and How
Feng Li, Xiaohe Wu, Wangmeng Zuo, David Zhang, Lei Zhang
https://arxiv.org/abs/1905.06648v1
https://github.com/lifeng9472/Removing_cosine_window_from_CF_trackers
显著性 | 利用眼动数据进行显著性建模
Leverage eye-movement data for saliency modeling: Invariance Analysis and a Robust New Model
Zhaohui Che, Ali Borji, Guangtao Zhai, Xiongkuo Min, Guodong Guo, Patrick Le Callet
https://arxiv.org/abs/1905.06803v1
https://github.com/CZHQuality/Sal-CFS-GAN
CVPR 2019
3D人脸建模 | 使用单图像进行3D人脸形状回归
Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision
Soubhik Sanyal, Timo Bolkart, Haiwen Feng, Michael J. Black
https://arxiv.org/abs/1905.06817v1
http://ringnet.is.tuebingen.mpg.de/
ICML 2019
对抗攻击 | 通过有效的组合优化进行简单的黑盒对抗性攻击
Parsimonious Black-Box Adversarial Attacks via Efficient Combinatorial Optimization
Seungyong Moon, Gaon An, Hyun Oh Song
https://arxiv.org/abs/1905.06635v1
https://github.com/snu-mllab/parsimonious-blackbox-attack
ICML 2019
数据增广 | 增广策略学习
Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules
Daniel Ho, Eric Liang, Ion Stoica, Pieter Abbeel, Xi Chen
https://arxiv.org/abs/1905.05393v1
https://github.com/arcelien/pba
点云数据处理识别 | 使用局部空间注意力模型的点集特征学习
LSANet: Feature Learning on Point Sets by Local Spatial Attention
Lin-Zhuo Chen, Xuan-Yi Li, Deng-Ping Fan, Ming-Ming Cheng, Kai Wang, Shao-Ping Lu
https://arxiv.org/abs/1905.05442v1
https://github.com/LinZhuoChen/LSANet
姿态估计| 多视图3D人体姿态估计，构建可学习的三角测量
Learnable Triangulation of Human Pose
Karim Iskakov, Egor Burkov, Victor Lempitsky, Yury Malkov
https://arxiv.org/abs/1905.05754v1
https://saic-violet.github.io/learnable-triangulation
ICIP 2019
人脸解析 | 通过域适应方法进行弱监督的漫画人脸解析
Weakly-supervised Caricature Face Parsing through Domain Adaptation
Wenqing Chu, Wei-Chih Hung, Yi-Hsuan Tsai, Deng Cai, Ming-Hsuan Yang
https://arxiv.org/abs/1905.05091v1
https://github.com/ZJULearning/CariFaceParsing
手势识别 | 设计手势音素。生成更大规模的不同手势，并用CNN识别
Talking with Your Hands: Scaling Hand Gestures and Recognition with CNNs
Okan Köpüklü, Yao Rong, Gerhard Rigoll
https://arxiv.org/abs/1905.04225
https://www.mmk.ei.tum.de/shgd/
域适应 | 提出一种叫Virtual Mixup Training 的正则化方法，用于非监督学习的域适应
Virtual Mixup Training for Unsupervised Domain Adaptation
Xudong Mao, Yun Ma, Zhenguo Yang, Yangbin Chen, Qing Li
https://arxiv.org/abs/1905.04215
（代码将开源）
表情识别 | 区域注意力网络用于针对姿态变化和遮挡鲁棒的人脸表情识别
Region Attention Networks for Pose and Occlusion Robust Facial Expression Recognition
Kai Wang, Xiaojiang Peng, Jianfei Yang, Debin Meng, Yu Qiao
https://arxiv.org/abs/1905.04075
https://github.com/kaiwang960112/Challenge-condition-FER-dataset
CVPR 2019
超分辨率 | 使用传感器RAW数据进行超分辨率，比RGB图像获得更好的效果
Zoom To Learn, Learn To Zoom
Xuaner Cecilia Zhang, Qifeng Chen, Ren Ng, Vladlen Koltun
https://arxiv.org/abs/1905.05169v1
https://ceciliavision.github.io/project-pages/project-zoom.html