Title: "Learning from Humans: Enabling a Scalable and Efficient Future of Robotics"
Abstract: Humans have deep expertise on an incredible array of skill sets. Increasingly robots are being trained to assist in performing some of these tasks, however robot learning is often limited in two major ways. First, increasingly complex tasks limit the human's ability to explicitly define task success. Second, modern methods of learning by imitation require a prohibitively large number of samples. We show how new methods of learning from interventions can overcome these hurdles and unlock flexible methods for learning from all types of human interaction.
Bio: Jonathan Spencer is a final year PhD candidate at Princeton University. He is especially interested in techniques for learning from humans and modeling human behavior and his work focuses on imitation learning and robotics. He is very passionate about the potential that robotics and autonomous driving have to save lives and serve the elderly and other communities with less access to transportation. Jonathan is an avid runner, cyclist, and baker, and hopes that one day robots will finally take over the world so he can do more running, biking, and baking.
To obtain the Zoom link for this event, please see the event announcements from MLFL on the college email lists or contact Kalpesh Krishna.