Visual Imitation Enables Contextual Humanoid Control
This work addresses the challenge of scalable teaching for humanoid robots to operate in diverse real-world environments, representing a novel method rather than an incremental improvement.
The authors tackled the problem of teaching humanoid robots to perform contextual skills like climbing stairs and sitting on chairs by introducing VIDEOMIMIC, a pipeline that mines everyday human motion videos to reconstruct humans and environments, producing whole-body control policies; they demonstrated robust, repeatable control on real robots for tasks such as staircase ascents and descents from a single policy.
How can we teach humanoids to climb staircases and sit on chairs using the surrounding environment context? Arguably, the simplest way is to just show them-casually capture a human motion video and feed it to humanoids. We introduce VIDEOMIMIC, a real-to-sim-to-real pipeline that mines everyday videos, jointly reconstructs the humans and the environment, and produces whole-body control policies for humanoid robots that perform the corresponding skills. We demonstrate the results of our pipeline on real humanoid robots, showing robust, repeatable contextual control such as staircase ascents and descents, sitting and standing from chairs and benches, as well as other dynamic whole-body skills-all from a single policy, conditioned on the environment and global root commands. VIDEOMIMIC offers a scalable path towards teaching humanoids to operate in diverse real-world environments.