Perceptive Humanoid Parkour chains dynamic obstacle skills on Unitree G1
Evidence notes
- Perceptive Humanoid Parkour combines motion matching, RL expert policies, and a depth-based multi-skill student policy for dynamic obstacle traversal.
- The authors validate the framework on a real Unitree G1, including climbing obstacles up to 1.25 m and long-horizon multi-obstacle traversal.
- The release is one of the more substantial external G1 results for onboard-perception parkour and dynamic locomotion.
Company context
Develops quadruped and humanoid robots, with strength in dynamic locomotion, vertically integrated hardware, and relatively low-cost commercial deployment. Unitree is one of the clearest examples of a legged robotics company moving from research visibility into real productisation and broader market distribution.