近五年图像分割和人体姿态识别的一些经典必看论文

一梯度:卡内基梅隆大学机器人研究所以CPM为基础,从单人到多人的实时姿态识别研究。

  1. Convolutional Pose Machines (CVPR2016 开源)
  2. DeeperCut: A Deeper, Stronger, and Faster Multi-person Pose Estimation Model (ECCV2016 开源)
  3. OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields (CVPR2017 开源)

二梯度:FAIR、谷歌大脑、清华&华科大,多种金字塔网络用于目标检测和多人姿态识别
1.Feature Pyramid Networks for Object Detection (CVPR2017 开源)
2.Towards Accurate Multi-person Pose Estimation in the Wild (CVPR2017)
3.Cascaded Pyramid Network for Multi-Person Pose Estimation(CVPR2018 开源)

三梯度:上交大提出的开源AIphaPose系统,效果比Mask R-CNN要好,用于实时视频识别
1.RMPE: Regional Multi-person Pose Estimation (ICCV2017 开源)
2.CrowdPose: Efficient Crowded Scenes Pose Estimation and A New Benchmark (CVPR2019 开源)

四梯度:多个2019年一些新方法,如自下而上的识别,3D识别等(比较杂乱)
1.PifPaf: Composite Fields for Human Pose Estimation (CVPR2019 开源)
2.PoseTrack: A Benchmark for Human Pose Estimation and Tracking (CVPR2018 开源)
3.3D human pose estimation in video with temporal convolutions and semi-supervised training (开源)
4.Exploiting Temporal Context for 3D Human Pose Estimation in the Wild (CVPR2019 开源)
5.Fast and Robust Multi-Person 3D Pose Estimation from Multiple Views (CVPR2019 开源)
6.PoseFix: Model-agnostic General Human Pose Refinement Network (CVPR2019 开源)

五梯度:图像分割、目标检测的鼻祖(卷积+deeplearning)
1.Rich feature hierarchies for accurate object detection and semantic segmentation (2013.11 开源)
2.Fast R-CNN (ICCV2015 开源)
3.Fully Convolutional Networks for Semantic Segmentation (CVPR2015 开源)
4.Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (NIPS2015 开源)

六梯度:图像分割进阶 (近三年的语义分割、实例分割、全景分割)
1.Mask R-CNN (CVPR2017 开源)
2.Rethinking the Faster R-CNN Architecture for Temporal Action Localization
3.Mask Scoring R-CNN (CVPR2018 开源)
4.Dual Attention Network for Scene Segmentation (CVPR2019 开源)
5.DFANet: Deep Feature Aggregation for Real-Time Semantic Segmentation (CVPR2019 开源)
6.Attention-guided Unified Network for Panoptic Segmentation (CVPR2019)

以上论文都是本人看过的一些论文,如果时间允许,也将会详细介绍每一篇的内容。
所有论文都可以在百度找到对应的原文和一些博客的解读,源码链接也在原文中或者解读博客中给出。
如有其他疑问请评论,博主看到会第一时间回复。
最后,希望大家在deep Learning这条路上越走越开心,✌!

你可能感兴趣的:(CVPR论文之人体姿态估计,CVPR论文之图像分割,deepinglearning)