We present a system for learning generalizable hand-object tracking controllers purely from synthetic data, without requiring any human demonstrations. Our approach makes two key contributions: (1) HOP, a Hand-Object Planner, which can synthesize diverse hand-object trajectories; and (2) HOT, a Hand-Object Tracker that bridges synthetic-to-physical transfer through reinforcement learning and interaction imitation learning, delivering a generalizable controller conditioned on target hand-object states. Our method extends to diverse object shapes and hand morphologies. Through extensive evaluations, we show that our approach enables dexterous hands to track challenging, long-horizon sequences including object re-arrangement and agile in-hand reorientation. These results represent a significant step toward scalable foundation controllers for manipulation that can learn entirely from synthetic data, breaking the data bottleneck that has long constrained progress in dexterous manipulation.
@article{arxiv.2512.19583,
title = {Learning Generalizable Hand-Object Tracking from Synthetic Demonstrations},
author = {Yinhuai Wang and Runyi Yu and Hok Wai Tsui and Xiaoyi Lin and Hui Zhang and Qihan Zhao and Ke Fan and Miao Li and Jie Song and Jingbo Wang and Qifeng Chen and Ping Tan},
journal= {arXiv preprint arXiv:2512.19583},
year = {2025}
}