Tencent Games Motus Debuts AI Pipeline for 3D Animation

The Motus platform bridges the gap between disparate animation tasks by applying foundation models to the entire character creation lifecycle. Rather than focusing solely on motion generation, the system analyzes character topology to produce skeletons and skinning weights in seconds. This approach addresses common technical hurdles like mesh penetration, foot sliding, and joint anomalies, which often negate the time-saving benefits of standalone generative tools.

Technical lead Zijiao Zeng detailed the framework during a presentation titled "Breathing Life into Geometry 2.0" at Gamescom Dev. The pipeline supports motion creation from diverse inputs, including text prompts, video, and keyframes, while enabling real-time responses for NPCs and digital humans. By automating the technical cleanup, the team intends to shift the focus of animators back toward creative expression and character design, a transition already tested across several Tencent game titles.