NVIDIA Research has presented Joints, Camera, Action! (JCA), a research system that jointly generates video, synchronized 3D motion, and camera alignment for a supplied rig of arbitrary topology. The project is described in a 2026 paper by Noam Rotstein and collaborators.
JCA combines a rig-conditioned motion model with a pretrained video generator. A graph encoder represents the skeleton, while bidirectional attention bridges connect the motion and video branches. The system can use a rest-pose rig, text prompt, and reference image for scene-aware animation; it also supports RGB keyframes for trajectory control.
The researchers say JCA can recover 3D rig motion from video and generate video from a rig-motion sequence, with or without a reference image. Its training material includes nearly 200,000 video-and-motion pairs, alongside motion-only data. The source reports comparative evaluations, but does not establish public availability of code, model weights, an API, pricing, or production performance.