传统的6d位姿估计fangfa1_【6D位姿估计】开源项目（持续更新）

2020

• PVN3D

[CVPR 2020] PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimation.

[ arXiv, Code]

PVN3Dhttps://www.zhihu.com/video/1232768319741755392

• 6-PACK

[ICRA 2020] 6-PACK: Category-level 6D Pose Tracker with Anchor-Based Keypoints.

[Project, arXiv, Code]

6-PACK

2019

• DenseFusion

[CVPR 2019] DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion.

[ Project , arXiv, Code ]

DenseFusionhttps://www.zhihu.com/video/1183137845279178752

• NOCS

[CVPR 2019] Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation.

[ Project, arXiv, Code]

NOCShttps://www.zhihu.com/video/1183138656461701120

• PVNet

[CVPR 2019] PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation.

[ Project, arXiv, Code]

PVNethttps://www.zhihu.com/video/1183138854776827904

论文作者的知乎帖：

彭思达：浙大CAD&CG实验室提出PVNet，实时且效果超群，已开源​zhuanlan.zhihu.com

• Pix2Pose

[ICCV 2019] Pix2Pose: Pixel-Wise Coordinate Regression of Objects for 6D Pose Estimation.

[ arXiv, Code]

Pix2Posehttps://www.zhihu.com/video/1233157174483632128

• MTTM

[ICRA 2019] Multi-Task Template Matching for Object Detection, Segmentation and Pose Estimation Using Depth Images.

[ Paper ]

MTTMhttps://www.zhihu.com/video/1233332378149326848

2018

• PoseCNN

[RSS 2018] PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes.

[ Project, arXiv, Code, YCB-Video Datasets Toolbox ]

PoseCNNhttps://www.zhihu.com/video/1183140676269801472

YCB-video 数据集简介：

丁洪凯：位姿估计数据集 YCB-Video Dataset​zhuanlan.zhihu.com

• DOPE

[CoRL 2018] Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects.

[ Project, arXiv, Code ]

DOPEhttps://www.zhihu.com/video/1183139670938075136

• DeepIM

[ECCV 2018] DeepIM: Deep Iterative Matching for 6D Pose Estimation.

[ Project, arXiv, Code]

DeepIM: Deep Iterative Matching for 6D Pose Estimation_哔哩哔哩 (゜-゜)つロ 干杯~-bilibili​www.bilibili.com

DeepIM: Deep Iterative Matching for 6D Pose Estimation (ECCV 2018)_哔哩哔哩 (゜-゜)つロ 干杯~-bilibili​www.bilibili.com

2017

• SSD-6D

[ICCV 2017] SSD-6D: Making RGB-based 3D detection and 6D pose estimation great again.

[ arXiv, Code ]

SSD-6D: Making RGB-Based 3D Detection and 6D Pose Estimation Great Again._哔哩哔哩 (゜-゜)つロ 干杯~-bilibili​www.bilibili.com

SSD6D - Supplementary Video_哔哩哔哩 (゜-゜)つロ 干杯~-bilibili​www.bilibili.com

• BB8

[ICCV 2017] BB8: A Scalable, Accurate, Robust to Partial Occlusion Method for Predicting the 3D Poses of Challenging Objects without Using Depth.

[ arXiv ]

经典的 Linemod

[1]. Multimodal Templates for Real-Time Detection of Texture-less Objects in Heavily Cluttered Scenes, IEEE International Conference on Computer Vision (ICCV), 2011.

[2]. Gradient Response Maps for Real-Time Detection of Texture-Less Objects. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2012.

[3]. Model Based Training, Detection and Pose Estimation of Texture-Less 3D Objects in Heavily Cluttered Scenes. [PDF ]

作者 Stefan Hinterstoisser 的个人网站：

http://campar.in.tum.de/Main/StefanHinterstoisser​campar.in.tum.de

OpenCV 实现：集成在 contrib 模块中

https://github.com/opencv/opencv_contrib/blob/master/modules/rgbd/src/linemod.cpp​github.com

ROS 实现：集成在 Object Recognition Kitchen (ORK) 库中，是基于 OpenCV Linemod 实现

Object Recognition Kitchen​wg-perception.github.io ROS kinetic + Realsens D435i + ORK + LINEMOD 物体识别​www.cnblogs.com

PPF

Model globally, match locally: Efficient and robust 3D object recognition. CVPR, 2010.

丁洪凯：【6D位姿估计】Point Pair Feature (PPF)​zhuanlan.zhihu.com

Benchmark

• BOP: Benchmark for 6D Object Pose Estimation

[ Project, Paper, BOP_Toolkit ]

部分数据集：