LEGS: Fine-Tuning Teleop-Free VLAs for Humanoid Loco-manipulation in an Embodied
Overview
This research presents LEGS, a hybrid simulator that combines a mesh foreground with a photorealistic 3D Gaussian Splatting background to train vision-language-action (VLA) policies for humanoid loco-manipulation without human teleoperation. The method uses a procedural motion-primitive generator to synthesize labeled demonstrations and a two-stage color calibration to align renders with the deployment camera. Policies trained on LEGS data match or exceed those trained on human teleoperation demos and outperform mesh-only baselines. The approach enables cost-effective re-rendering of demonstrations for new scenes and objects.
From arXiv Robotics — research abstracts
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · Unitree G1 humanoid robot: 1 other · Basis not reported
“-and-place tasks of increasing whole-body difficulty and three VLA backbones ($\psi_0$, $\pi_{0.5}$, GR00T N1.6), a policy trained purely on LEGS data matches or exceeds one trained on human teleoperation demos on every experiment. It also outperforms a mesh-only baseline trained on the same motions without the calibrated 3DGS frontend. Because motion is recorded independently of appearance, the same demonstrations c”
Exact source · revision 1Source owner not reported
Reported quantity · pick-and-place tasks: 3 other · Basis not reported
“-and-place tasks of increasing whole-body difficulty and three VLA backbones ($\psi_0$, $\pi_{0.5}$, GR00T N1.6), a policy trained purely on LEGS data matches or exceeds one trained on human teleoperation demos on every experiment. It also outperforms a mesh-only baseline trained on the same motions without the calibrated 3DGS frontend. Because motion is recorded independently of appearance, the same demonstrations c”
Exact source · revision 1Source owner not reported
Reported quantity · VLAs backbones: 3 other · Basis not reported
“-and-place tasks of increasing whole-body difficulty and three VLA backbones ($\psi_0$, $\pi_{0.5}$, GR00T N1.6), a policy trained purely on LEGS data matches or exceeds one trained on human teleoperation demos on every experiment. It also outperforms a mesh-only baseline trained on the same motions without the calibrated 3DGS frontend. Because motion is recorded independently of appearance, the same demonstrations c”
Exact source · revision 1Source owner not reported
Reported quantity · policy trained on LEGS data: 1 other · Basis not reported
“-and-place tasks of increasing whole-body difficulty and three VLA backbones ($\psi_0$, $\pi_{0.5}$, GR00T N1.6), a policy trained purely on LEGS data matches or exceeds one trained on human teleoperation demos on every experiment. It also outperforms a mesh-only baseline trained on the same motions without the calibrated 3DGS frontend. Because motion is recorded independently of appearance, the same demonstrations c”
Exact source · revision 1Source owner not reported
Artifact availability · code: available
“an be re-rendered for a target scene and objects at over 15x lower cost than teleoperation. When both the objects and the scene change, a policy trained on LEGS-AUG data re-rendered for that target maintains task success, while the teleop baseline fails entirely. Code and videos are available at https://hojjunekim.github.io/legs/.”
Exact source · revision 1Source owner not reported
Evidence dependency changes (2)
- 2026-10-30T22:57:10.423Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T22:56:56.548Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalLEGS: Fine-Tuning Teleop-Free VLAs for Humanoid Loco-manipulation in an Embodied Gaussian Splatting World
This research presents LEGS, a hybrid simulator that combines a mesh foreground with a photorealistic 3D Gaussian Splatting background to train vision-language-action (VLA) policies for humanoid loco-manipulation without human teleoperation. The method uses a procedural motion-primitive generator to synthesize labeled demonstrations and a two-stage color calibration to align renders with the deployment camera. Policies trained on LEGS data match or exceed those trained on human teleoperation demos and outperform mesh-only baselines. The approach enables cost-effective re-rendering of demonstrations for new scenes and objects.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).