EPFL (Mathis Group) · Musculoskeletal Locomotion

TERRA

Terrain-Aware Reconstruction, Retargeting and Control, from EPFL's Mathis group (the team behind MuscleMimic and Kinesis), teaches simulated muscle-driven human bodies to walk over terrain that is not flat. It infers the supporting surfaces from motion-capture data, applies biomechanical constraints when transferring the motion, and trains controllers on the result. The locomotion library totals 11.3 hours, 9.4 of them used for training, and the paper shows steep ramps, stairs, seats, separate platforms, narrow beams and uneven ground. Paper arXiv 2609.38653 (29 September); Apache-2.0 code on GitHub, updated through 9 October.

Open the source ↗

Similar Agents & Automation releases

  • SceneBloom

    Yuhan Liu, Lianhui Qin, Shenji Wan · Indoor scene generation · Week 45, 2026

  • MotionBricks

    NVIDIA (with ETH Zürich, Simon Fraser University and the University of Texas at Austin) · Real-time motion generation · Week 45, 2026

  • RoboParty RP1

    RoboParty · Open Source Humanoid · Week 43, 2026

  • FLUX 3 Action

    Black Forest Labs · Robot Action Model · Week 40, 2026

  • GPT-Policy

    Cheng, Yi, Fang and 12 co-authors · In-Context Robot Learning · Week 39, 2026

  • UnifoLM-WLA-1.0

    Unitree Robotics · Humanoid World Model VLA · Week 37, 2026

Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.