Enforcing Human-like Kinematics in Dexterous Piano Playing via Adversarial Posture Regularization
For researchers in dexterous manipulation and animation, APR provides a practical method to improve naturalness of learned policies without expensive expert demonstrations.
Reinforcement learning for dexterous piano playing often produces unnatural postures; the proposed Adversarial Posture Regularization (APR) uses a small amount of casual human playing data to enforce human-like kinematics, achieving significantly better human-likeness metrics (cPSI, BSE, FAC) and visual quality compared to prior methods.
Reinforcement learning can train bimanual dexterous hands to play piano in physics simulation with high note accuracy, but for high-DoF dexterous hands, relying solely on task rewards or IK inversion often leads to unnatural postures and joint overextension. We propose \textit{Adversarial Posture Regularization (APR)}. It avoids expensive, song-aligned expert demonstration data and instead uses a small amount of casual human playing data. By matching the distribution of the posture of the policy with the human prior through an adversarial objective, APR encourages more human-like hand shapes. Meanwhile, we collect and release unstructured hand motion data of piano playing using a consumer-grade Meta Quest 3, and retarget the key motion information to the Shadow Hand. Finally, we achieve significantly better performance than prior methods on all three human-likeness metrics (cPSI, BSE, and FAC) as well as in visual quality. Project repository: https://github.com/APRProject/APRPianist.