· Contributors · Organizations · Search
Learning to Get Up
Research & Education
Full Conference Supporter
Exhibitor Additional Full Conference
Exhibitor Full Conference
This session WILL NOT be recorded.
LocationEast Building, Ballroom A/B
DescriptionWe develop a method for generating natural and diverse human get-up behaviors using deep reinforcement learning. The learned control policies enable characters to get up from any initial fallen pose, either slowly or quickly. We visualize the solution modes learned by the controllers from various initial states.