Edge Grasp Network: A Graph-Based SE(3)-invariant Approach to Grasp Detection
About
Given point cloud input, the problem of 6-DoF grasp pose detection is to identify a set of hand poses in SE(3) from which an object can be successfully grasped. This important problem has many practical applications. Here we propose a novel method and neural network model that enables better grasp success rates relative to what is available in the literature. The method takes standard point cloud data as input and works well with single-view point clouds observed from arbitrary viewing directions.
Haojie Huang, Dian Wang, Xupeng Zhu, Robin Walters, Robert Platt• 2022
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Clutter removal | Packed scenes single-view, fixed camera, gamma noise | GSR60.6 | 16 | |
| Clutter removal | Pile scenes single-view, fixed camera, gamma noise | GSR55 | 16 | |
| Clutter removal | Pile single-view, random camera pose, Gaussian noise | GSR92 | 10 | |
| Clutter removal | Packed single-view, random camera pose, Gaussian noise | GSR92.5 | 10 | |
| Clutter removal | Real-world Adv | GSR79.5 | 7 | |
| Clutter removal | Pile Real-world | GSR (%)67.7 | 7 | |
| Clutter removal | Real-world Packed | GSR73.4 | 7 |
Showing 7 of 7 rows