Sagar Parekh

Papers

Total Citations

H-Index

About

Sagar Parekh is a robotics researcher whose work sits at the intersection of visual imitation learning, assistive robotics, and human-robot interaction. His research tackles one of the central challenges in modern robotics: enabling machines to learn complex, generalizable behaviors with minimal supervision or demonstration overhead. Parekh is perhaps best recognized for his development of VIEW (Visual Imitation Learning with Waypoints), a framework that addresses the inherent difficulties of translating high-dimensional video observations into actionable robot manipulation policies. By introducing waypoint-based representations, his approach offers a more tractable path for robots learning from video demonstrations — a contribution that has already attracted notable attention, accumulating 10 citations shortly after its 2025 publication. His earlier work on learning latent actions without human demonstrations (2022, 7 citations) reflects a strong commitment to accessibility and inclusivity in robotics. By enabling assistive robots to acquire meaningful action mappings from disabled users' low-dimensional joystick inputs — without requiring non-disabled human demonstrations — Parekh directly addresses real-world barriers in assistive technology deployment. Across his body of work, Parekh consistently pushes toward robot learning systems that are both practically deployable and socially meaningful, making him a researcher worth watching as the field continues to mature.