Runway, a pioneer in AI-powered creative tools, has launched its next-generation motion capture model, Act-Two, bringing a remarkable breakthrough to digital content creation. Act-Two enables animators to transfer the full range of body, facial, head, and hand movements to any character using only a driving performance video and a reference image or video—eliminating the need for expensive, specialized hardware.
With Act-Two, users simply upload a video of a person performing the desired scene along with the image or video of the character they want to animate. The model then accurately transfers the nuances of movement, expression, speech, and gesture to bring the character to life in a realistic manner. When using still images, Act-Two automatically adds environmental motion and subtle camera movements for a more natural result. Gesture control settings allow specific body language and poses from the driving performance to be faithfully mirrored by the animated character.
This makes Act-Two a powerful and accessible tool for creators across backgrounds and skill levels. It supports impressive outcomes with both human and non-human characters, working across varied styles and camera angles. Accessible via Runway's web platform, the model operates on a credits-based system (five credits per second) and supports outputs in multiple resolutions, including 1280x720 and 720x1280. Runway's own "restyle" feature further opens up creative possibilities by enabling unique visual styles for characters.
As a major leap in AI filmmaking, Act-Two democratizes cutting-edge motion capture, offering unprecedented opportunities for filmmakers, advertisers, game developers, and educators—ushering in a new era of digital storytelling.