Watch Your Step: Anticipated Perception for Humanoid Locomotion
Perceptive locomotion policies typically observe a dense robot-centric elevation map, although only the terrain beneath and around the next footsteps informs the control decision.
ProofPaper ↗
Key points
- Learned encoders compress the whole visual input, leaving the policy to work out which inputs matter while learning the baseline locomotion task.
- A lightweight foothold prediction model (FPM) predicts future footholds from proprioception and the velocity command alone, and we sample the elevation map only in small patches around these predicted footholds.
- We train the prediction model alongside the locomotion policy, periodically retraining it on new policy rollouts as the robot's gait matures.
- This allows the perception to anticipate where the policy will step, while the policy learns to leverage terrain information around those steps.
Sources (1)
- [1]Watch Your Step: Anticipated Perception for Humanoid LocomotionarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 21, 11:52 AM
Perceptive locomotion policies typically observe a dense robot-centric elevation map, although only the terrain beneath and around the next footsteps informs the control decision.
Learned encoders compress the whole visual input, leaving the policy to work out which inputs matter while learning the baseline locomotion task.
Extractive summary: sentences quoted from the sources.