Motion capture session for AI-assisted character animation
Motion capture recordings have laid the groundwork for training the AI framework, AnimHost, a key component of the MAX-R initiative at Filmakademie Baden- Württemberg.
Within the scope of the project, the R&D team at the Animationsinstitut is developing a toolset that utilises motion capture data to replicate an actor's movements in a given role. Through this technology, AnimHost "learns" to understand the nuances of bipedal movement, enabling the interactive reproduction of these movements and allowing virtual characters to mimic the performance style of real actors for various applications such as previsualisation and short films. While these open-source AI tools under development are not aimed at replacing artists, they are specifically designed to empower creatives by improving established workflows.
At present, the team explores the AI-driven animation on the little robot MAX, teaching him the intricacies of jumping, walking, running and beyond. If you’re interested in tracking his advancements, we invite you to watch the linked video for an up-close look at MAX's dynamic capabilities in action.
