Animate static human images into expressive, lifelike videos using DreamActor-M1. Powered by diffusion transformers with hybrid guidance, it enables fine control of facial expressions, body motion.
DreamActor-M1 is a diffusion transformer (DiT)-based framework for animating human images. It takes a single reference image (portrait or full body) plus a driving video or motion signals and generates a realistic animated video of the person performing that motion.