yzqin / dexpoint-release

DexPoint: Generalizable Point Cloud Reinforcement Learning for Sim-to-Real Dexterous Manipulation, CoRL 2022

https://yzqin.github.io/dexpoint/

yzqin/dexpoint-release Issues

Reproducible issue: cannot converge to a valid grasp policy
Closed 7 months ago2
Regarding Loading a custom object and the replication of training code
Closed 7 months ago2
Training issue: Should get_observation return array instead of dict？
Closed 7 months ago1
Is the pre-trained pointnet used in this paper equal to the one in dexart？
Updated 7 months ago2
Request for scenario files or methods for creating them
Closed 9 months ago4
How to avoid penetration between dexterous hands and objects？
Closed 10 months ago4
How to draw the visual point cloud simulation like in Project?
Updated a year ago2