Quick Training of RL Models for HSK Grasping: Experience Sharing

Due to certain training parameters and model designs involving company confidentiality, they will not be disclosed. The following only includes methods provided by past literature.

MaterialPermalink

Slides

GithubPermalink

Add two new environment:

• Image Shot: PandaPickAndPlaceShot-v3
• Oriented Bounding Box (OBB): PandaPickAndPlace6d-v3

Code

ReferencesPermalink

1. A. Goyal, J. Xu, Y. Guo, V. Blukis, Y.-W. Chao, D. Fox, (2023), RVT: Robotic View Transformer for 3D Object Manipulation, PMLR.
2. Y. Geng, B. An, H. Geng, Y. Chen; Y. Yang, H. Dong, (2023), RLAfford: End-to-End Affordance Learning for Robotic Manipulation, ICRA.
3. W. Huang, C. Wang, R. Zhang, Y. Li, J. Wu, Li F.-F., (2023), VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Models, CoRL.

AcknowledgementsPermalink

I sincerely thank my supervisor Chen and Mr. Jiang for the guidance and encouragement.