ReDexT

Learning a Generalizable Residual Policy for Dexterous Retargeting

Jin-Chuan Shi^* · Yangjinhui Xu^* · Liyang Li · Muzhi Zhu · Jiadong Hong · Yue Hu · Hao Chen · Chunhua Shen

State Key Lab of CAD & CG, Zhejiang University
^*Equal contribution

Project Page  ·  Paper

ReDexT learns shared residual feedback for dexterous retargeting. A trained policy corrects inverse-kinematics commands to execute new human hand-object trajectories with frozen weights, and supports optional local or shared adaptation for improved tracking.

Code release

The implementation is being prepared for release. This repository will host the code and instructions for training and evaluation.

The project page contains the paper, quantitative results, and simulation videos across five robot hands.

Citation

@misc{shi2026redext,
  title = {ReDexT: Learning a Generalizable Residual Policy for Dexterous Retargeting},
  author = {Shi, Jin-Chuan and Xu, Yangjinhui and Li, Liyang and Zhu, Muzhi and Hong, Jiadong and Hu, Yue and Chen, Hao and Shen, Chunhua},
  year = {2026},
  url = {https://aim-uofa.github.io/ReDexT/}
}