DexGraspNet 2.0
PKU-EPIC
Generative dexterous grasping tools for large synthetic cluttered scenes.
What you can explore
Explore the released learning pipeline, object representations and grasp evaluation.
Before you use it
Separate from the original DexGraspNet dataset record. Requires the documented CUDA/Isaac Gym stack and separately downloaded assets.
Original contributors
Jialiang Zhang, Haoran Liu and collaborators
License / access: CC BY-NC 4.0 · work and dataset; check dependency terms separately
Editorial reference · checked 2026-09-22