DONIC: Generalist Dexterous Bimanual Tracking Control from Human Demonstrations

DONIC is presented as a single generalist policy that controls two dexterous robot hands to track bimanual references derived from human demonstrations. It is trained with PPO plus behavioral cloning, using recycled successful executions, an object-centered per-hand reward, and Clip-Centered Advantages. The authors also introduce BiHOI, a multi-view dataset of 1,575 bimanual sequences over 229 objects. In simulation, without fine-tuning, the policy reports 86.79% success on unseen trajectories of training objects and 57.32% on held-out objects, which the source says exceed bimanual adaptations of DexTrack and Dexplore.

Why it matters

Bimanual dexterous tracking has been less studied than single-hand control, and a single policy generalizing to unseen objects without fine-tuning would reduce per-task engineering for two-hand robot manipulation.

Read with care

Reported results are from simulation and the baselines are the authors' own bimanual adaptations; source claims not independently checked.