ETH Zurich's ADAPT runs diffusion action priors for real-time language-driven humanoid control at 50Hz
breadli428 · x · 2026-11-03
ETH Zurich researchers present ADAPT (CoRL 2026), an end-to-end closed-loop framework for text-conditioned humanoid whole-body control. Unlike dominant text-to-motion pipelines that generate kinematic motions for a separate tracker, ADAPT learns a diffusion-based action prior from text-labeled humanoid trajectories, mapping changing language commands directly to dynamically feasible actions.
- Runs real-time policy inference at 50Hz with smooth online prompt switching, deployed on a real Unitree G1
- A lightweight residual RL policy trained on top of the frozen diffusion controller improves long-horizon robustness
- The same diffusion prior is reusable as a steerable motion prior for downstream tasks like goal-reaching with controllable motion styles
Related event: ETH Zurich's ADAPT Enables Real-Time Language Control of Humanoid Robots(3 posts)→
More from Embodied
- ADAPT architecture breakdown: diffusion Transformer plus residual RL in four stages — breadli428 · 2026-11-03
- ADAPT: ETH Zurich's 50Hz end-to-end text-driven humanoid control with diffusion priors — breadli428 · 2026-11-03
- Xiaomi open-sources Robotics-1, a robot foundation model trained on 100K hours of trajectories — CyberRobooo · 2026-11-03
- Factory photos show desktop robot modules in volume assembly, project hits production stage — victormustar · 2026-11-03
- Machenta demos instant robot learning from passersby at London King's Cross — Ed__Johns · 2026-11-03
- PRIMA animal 3D mesh recovery accepted to WACV2027 with 9K pseudo-3D dataset — TrackingActions · 2026-11-03