Neural State Machine for Scene Interaction
A Neural State Machine has been presented at SIGGRAPH Asia 2019, featuring a data-driven deep learning framework that can handle complex animations. Transitions such as walking to sit on a chair, lift a box, or pass an openning a door are shown in the video down below. The system is able to learn character-scene interactions from motion capture data, and produces high-quality animations from simple control commands. Next-Gen animation?