RoboPianist: Dexterous Piano Playing with Deep Reinforcement Learning
This work addresses the problem of achieving advanced dexterity in robotics for applications like piano playing, representing an incremental step by applying existing methods to a new, complex domain.
The researchers tackled the challenge of replicating human-like dexterity in robot hands by using deep reinforcement learning to enable simulated anthropomorphic hands to learn 150 piano pieces, addressing high-dimensional control and complex finger coordination.
Replicating human-like dexterity in robot hands represents one of the largest open problems in robotics. Reinforcement learning is a promising approach that has achieved impressive progress in the last few years; however, the class of problems it has typically addressed corresponds to a rather narrow definition of dexterity as compared to human capabilities. To address this gap, we investigate piano-playing, a skill that challenges even the human limits of dexterity, as a means to test high-dimensional control, and which requires high spatial and temporal precision, and complex finger coordination and planning. We introduce RoboPianist, a system that enables simulated anthropomorphic hands to learn an extensive repertoire of 150 piano pieces where traditional model-based optimization struggles. We additionally introduce an open-sourced environment, benchmark of tasks, interpretable evaluation metrics, and open challenges for future study. Our website featuring videos, code, and datasets is available at https://kzakka.com/robopianist/