Yinlong Dai

Papers

Total Citations

H-Index

About

Yinlong Dai is a robotics researcher advancing the frontier of dexterous manipulation, with a focus on enabling multi-fingered robot hands to perform complex, contact-rich tasks. His work bridges the gap between human dexterity and robotic capability through three key contributions: integrating tactile sensing with visual cues for precise manipulation, learning from human demonstration videos without teleoperation, and democratizing robotics research with open-source tools. His most-cited paper, "See to Touch: Learning Tactile Dexterity through Visual Incentives" (2024, 18 citations), demonstrates how combining tactile feedback with visual information can significantly improve a robot's ability to reason about object spatial configurations—a critical challenge in the field. In "Bridging the Human to Robot Dexterity Gap Through Object-Oriented Rewards" (2025), he tackles the difficult problem of training multi-fingered hands directly from human video data, bypassing the need for expensive teleoperation setups. Additionally, his work on "OPEN TEACH" (2024) provides an open-source teleoperation system that makes advanced manipulation research accessible to the broader community. Dai's research is particularly notable for its practical approach to solving real-world manipulation challenges, with potential applications spanning logistics, home robotics, and beyond.