UniGraspTransformer
microsoft
Transformer policies distilled from specialist grasping policies to handle varied objects.
What you can explore
Study the state-based and vision-based training structure and the available specialist policy examples.
Before you use it
The README still marks some universal policy releases as coming soon. Follow the specific Isaac Gym version and asset setup.
Original contributors
Microsoft Research Asia / UniGraspTransformer authors
License / access: Code: MIT · data, models and dependencies may have separate terms
Keep exploring
UniDexGrasp++
A generalist-and-specialist training approach for dexterous grasping across object geometries.
DexGraspNet
Synthetic dexterous grasp configurations generated for a Shadow Hand and validated in simulation.
Isaac Lab
GPU-oriented robot-learning framework with articulated-robot manipulation environments.
Editorial reference · checked 2026-09-22