Data Flywheel for Humanoid Robotics
How do we generate enough useful data for humanoid robots to keep learning and improving?
Dr. Fan Shi will explore how physical simulation, digital twins, and reward modelling can help scale expensive real-world data with large-scale synthetic data, addressing one of the key challenges in humanoid robotics: generating sufficient high-quality training data for increasingly complex tasks.
Drawing from his work in robot learning for humanoid manipulation and locomotion, he'll discuss how simulation can support the development of embodied AI, and how learning in simulated environments can translate to robots operating in the physical world.
He'll also touch on what it takes to move these advances beyond research, and the opportunities and challenges in translating developments in humanoid robotics into real-world applications.
More About the Speaker
Fan Shi is an Assistant Professor in the Department of Electrical and Computer Engineering at NUS, where he holds the NUS Presidential Young Professorship. His research focuses on AI for robotics, with particular interests in physical simulation and robot learning. He has received several international recognitions, including awards and support from leading organizations such as the NVIDIA Academic Grant Program, Google Research Funding, and Swiss AI Initiative.
Before joining NUS, he was a Postdoctoral Researcher at the AI Center, ETH Zurich, working with Robotic Systems Lab and the Computational Robotics Lab. He earned his Ph.D. and M.S. degrees at the JSK Lab, University of Tokyo, and his B.S. degree at Yuanpei College, Peking University.
Lab page link (https://nus-hcrl.github.io/)
More About the Series
Singapore Embodied AI Colloquium (SEAIC) is a focused technical research series built around researcher-driven discussions and sharing on frontier topics in embodied AI, robot learning, and physical AI.
Get involved: Learn more about Lorong AI | Speaker Sign-up | WhatsApp Community | LinkedIn | X