DONIC: Generalist Dexterous Bimanual Tracking Control from Human Demonstrations

Despite recent progress in dexterous reference tracking, existing models focus primarily on single-hand manipulation and a limited range of interactions, leaving the bimanual dexterous behaviors common in real life unaddressed. The scarcity of contact-rich bimanual manipulation data and the increased complexity of coordinating contacts make developing a generalist bimanual tracker particularly challenging. Therefore, we present DONIC, a generalist bimanual tracker jointly trained across operation categories and human-motion sources, together with BiHOI, a multi-view dataset of 1,575 sequences spanning 229 objects and emphasizing sustained bimanual interaction. Reconstructed interactions are retargeted into robot references. DONIC combines proximal policy optimization with behavior cloning in a feedback loop that recycles verified successful executions into training demonstrations. An object-centered per-hand reward maintains object-motion feedback while supervising each hand according to its reference contact role. Clip-Centered Advantages remove reference-dependent advantage offsets while preserving within-trajectory differences. Together, these designs enable one policy to learn coordinated manipulation across heterogeneous references. Without fine-tuning, a single checkpoint achieves 86.79% success on unseen trajectories of training objects and 57.32% on held-out objects, exceeding the stronger of two bimanual tracking baselines by 77.36 and 45.73 percentage points.