Date: September 16, 2026
Speaker: Homanga Bharadhwaj, Assistant Professor, Johns Hopkins University
Title: Observational Learning for Manipulation via Visual Imitation of Humans
Host: Kuan Fang
Abstract: Robots learning to interact with the world by directly observing humans go about their daily life, has been a dream in AI for decades. In this talk, I will discuss recent advances in learning such robotic manipulation from large-scale web videos, and also enabling robots to follow in-context human video demonstrations one-shot / few-shot in novel scenarios. Finally, I'll conclude with a reality check about how far we are from the dream, and discuss open challenges.
Bio: Homanga Bharadhwaj is an Assistant Professor of Computer Science at Johns Hopkins University. He completed his Ph.D. from Carnegie Mellon in 2025, and before that an MSc from the University of Toronto, and a BTech from IIT Kanpur. Homanga was named a Future Leader in Robotics and AI by the University of Maryland in 2025, and his research has received a Best Dissertation Honorable Mention at CMU Robotics Institute, the Best Robot Manipulation Paper Finalist at ICRA '24, the Best Conference Paper Award at ICRA '24, and the Outstanding Presentation award at NeurIPS '23 Robot Learning Workshop.
