SOURCE-LINKED INTELLIGENCE

Sample, Simulate, Select: Physics-in-the-Loop Text-to-Motion for Humanoids Without Training

arXiv · AI, language, vision and robotics · article · Sep 22, 2026 · UTC

Text-to-motion models generate plausible human motion but do not model a robot's dynamics; whole-body tracking controllers execute robot references reliably but cannot replan an infeasible one. Recent language-to-humanoid systems bridge this gap by training. We measure how much of the gap closes with no training at all, by putting the deployment controller itself in the loop. Sample-simulate-select (S$^3$) draws $N$ motions per prompt from a frozen text-to-motion model, retargets each to a Unitree G1 by direction-matching inverse kinematics, rolls all of them out under full rigid-body dynamics

Read original source ↗ Open in workspace

recordType

paper

region

Global