ManipTrans: Efficient Dexterous Bimanual Manipulation Transfer via Residual Learning

This model is described in the paper ManipTrans: Efficient Dexterous Bimanual Manipulation Transfer via Residual Learning. It's a two-stage method for efficiently transferring human bimanual skills to dexterous robotic hands in simulation. The model first pre-trains a generalist trajectory imitator and then fine-tunes a specific residual module.

For code and usage instructions please see the project's Github repository: ManipTrans.