“Animating characters is a difficult task when it comes to interacting with objects and the environment. In this research, the Neural State Machine uses a data-driven deep learning framework that can handle such animations. The system is able to learn character-scene interactions from motion capture data, and produces high-quality animations from simple control commands. The framework can be used for creating natural animations in games and films, and is the first of such frameworks to handle scene interaction tasks for data-driven character animation. The research is implemented in Unity and TensorFlow, and published under the ACM Transactions on Graphics / SIGGRAPH Asia 2019.”

Find the paper here.