GDC is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

View, browse and sort the ever-growing list of sessions by pass type, track, and format. With this Session Viewer, you can view GDC 2023 session details, speakers and share your favorites via social media. You will be able to build your schedule and access it during the show via export or with the Mobile App, once live. Sessions do fill up and seating is first come, first serve, so arrive early to sessions that you would like to attend.

Machine Learning Summit: Wobbledoll: ML powered self-balancing synthesis ragdoll

Nan Ma  (Senior Engine Developer, LightSpeed Studios)

Pass Type: All Access Pass, Summits Pass

Topic: Programming

Format: Lecture

Vault Recording: TBD

Audience Level: All

Tired of brain-dead ragdolls? Want your character to be more physically "alive"? In this talk the speakers will present how reinforcement learning could turn a ragdoll into a "Wobbledoll", a self-balanced ragdoll that can perform locomotion and fully interactable in a virtually simulated world. Speakers will cover the technical details, such as machine learning, physics simulation and animation pipeline integration, as well as discussing how to leverage AI powered system to improve the character motion quality, interactiveness and gameplay. AI training analysis and in-game profile results along with the tech demo will also be presented in this talk, to provide in-depth technical review of the performance for both offline training and real-time execution.


The audience will develop insight into how a comprehensive physical animation system could be created by leveraging machine learning techniques and applied to improve the gameplay.

Intended Audience

All levels of AI/Physics/Animation programmers and technical animators will probably gain the most value from this talk, although no prior knowledge is required.