OmniHuman-1 is an advanced AI framework by ByteDance that generates realistic human videos from a single image and motion signals, such as audio or video.
