Himanshu Singh

Galgotias University, University of California, Berkeley

Papers

Total Citations

H-Index

About

Himanshu Singh’s research bridges assistive technology and robot learning, with a focus on human-centered AI and dexterous manipulation. His early work, “Gesture Control Robot using Arduino” (2019, 5 citations), addresses a critical accessibility gap: with only 10% of the world’s Deaf population receiving education in sign language, Singh developed a gesture-controlled robot to enable communication between sign language users and those unfamiliar with it. This project highlights his commitment to inclusive technology. More recently, his 2025 paper “Hand-Object Interaction Pretraining from Videos” (4 citations) introduces a novel framework for learning general robot manipulation priors by extracting 3D hand-object interaction trajectories from in-the-wild videos. By lifting both human hands and manipulated objects into a shared 3D space, Singh’s approach generates sensorimotor robot trajectories without costly manual data collection. This work has the potential to accelerate robot learning for tasks like grasping and assembly. Though early in his career, Singh’s contributions span both socially impactful applications and cutting-edge robot learning, making him a promising voice in human-robot interaction and embodied AI.