# LEGS: Fine-Tuning Teleop-Free VLAs for Humanoid Loco-manipulation in an Embodied Gaussian Splatting World arXiv: https://arxiv.org/abs/2606.01458 · HTML: https://mlsanity.com/paper/2606.01458/ ## Key facts - Cited by 2 papers in the map (data through 14 Oct 2026). - Placed in Tactile, force-aware and humanoid VLA, one of 13 directions in Vision-language-action models. - Builds on 35 of its 37 references that are in the map. - First posted on arXiv 31 May 2026. ## Placed in - Vision-language-action models / [Tactile, force-aware and humanoid VLA](https://mlsanity.com/vla/tactile-force-aware-and-humanoid-vla/) ## Builds on (most cited first, 30 of 35) - [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023) - [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024) - [π0: A Vision-Language-Action Flow Model for General Robot Control](https://mlsanity.com/paper/2410.24164/) (2024) - [π0.5: a Vision-Language-Action Model with Open-World Generalization](https://mlsanity.com/paper/2504.16054/) (2025) - [Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World](https://mlsanity.com/paper/1703.06907/) (2017) - [GR00T N1: An Open Foundation Model for Generalist Humanoid Robots](https://mlsanity.com/paper/2503.14734/) (2025) - [Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots](https://mlsanity.com/paper/2402.10329/) (2024) - [SAM 3D: 3Dfy Anything in Images](https://mlsanity.com/paper/2511.16624/) (2025) - [Foundation Models in Robotics: Applications, Challenges, and the Future](https://mlsanity.com/paper/2312.07843/) (2023) - [OmniH2O: Universal and Dexterous Human-to-Humanoid Whole-Body Teleoperation and Learning](https://mlsanity.com/paper/2406.08858/) (2024) - [HumanPlus: Humanoid Shadowing and Imitation from Humans](https://mlsanity.com/paper/2406.10454/) (2024) - [TWIST: Teleoperated Whole-Body Imitation System](https://mlsanity.com/paper/2505.02833/) (2025) - [SONIC: Supersizing Motion Tracking for Natural Humanoid Whole-Body Control](https://mlsanity.com/paper/2511.07820/) (2025) - [HOMIE: Humanoid Loco-Manipulation with Isomorphic Exoskeleton Cockpit](https://mlsanity.com/paper/2502.13013/) (2025) - [Approximate Convex Decomposition for 3D Meshes with Collision-Aware Concavity and Tree Search](https://mlsanity.com/paper/2205.02961/) (2022) - [Mobile-TeleVision: Predictive Motion Priors for Humanoid Whole-Body Control](https://mlsanity.com/paper/2412.07773/) (2024) - [Opt2Skill: Imitating Dynamically-feasible Whole-Body Trajectories for Versatile Humanoid Loco-Manipulation](https://mlsanity.com/paper/2409.20514/) (2024) - [Splat-MOVER: Multi-Stage, Open-Vocabulary Robotic Manipulation via Editable Gaussian Splatting](https://mlsanity.com/paper/2405.04378/) (2024) - [WholeBodyVLA: Towards Unified Latent VLA for Whole-Body Loco-Manipulation Control](https://mlsanity.com/paper/2512.11047/) (2025) - [Psi0: An Open Foundation Model Towards Universal Humanoid Loco-Manipulation](https://mlsanity.com/paper/2603.12263/) (2026) - [VIRAL: Visual Sim-to-Real at Scale for Humanoid Loco-Manipulation](https://mlsanity.com/paper/2511.15200/) (2025) - [Viser: Imperative, Web-based 3D Visualization in Python](https://mlsanity.com/paper/2507.22885/) (2025) - [GaussGym: An open-source real-to-sim framework for learning locomotion from pixels](https://mlsanity.com/paper/2510.15352/) (2025) - [EgoHumanoid: Unlocking In-the-Wild Loco-Manipulation with Robot-Free Egocentric Demonstration](https://mlsanity.com/paper/2602.10106/) (2026) - [GSWorld: Closed-Loop Photo-Realistic Simulation Suite for Robotic Manipulation](https://mlsanity.com/paper/2510.20813/) (2025) - [SOUS VIDE: Cooking Visual Drone Navigation Policies in a Gaussian Splatting Vacuum](https://mlsanity.com/paper/2412.16346/) (2024) - [Opening the Sim-to-Real Door for Humanoid Pixel-to-Action Policy Transfer](https://mlsanity.com/paper/2512.01061/) (2025) - [GRaD-Nav++: Vision-Language Model Enabled Visual Drone Navigation with Gaussian Radiance Fields and Differentiable Dynamics](https://mlsanity.com/paper/2506.14009/) (2025) - [Humanoid Manipulation Interface: Humanoid Whole-Body Manipulation from Robot-Free Demonstrations](https://mlsanity.com/paper/2602.06643/) (2026) - [DemoHLM: From One Demonstration to Generalizable Humanoid Loco-Manipulation](https://mlsanity.com/paper/2510.11258/) (2025) ## Built on by (most cited first, 2 of 2) - [Workhorse: Learning Robust Whole-Body Humanoid Loco-Manipulation from Human Data](https://mlsanity.com/paper/2610.09117/) (2026) - [Splat2Act: From Captured Scenes to Personalized VLA Manipulation](https://mlsanity.com/paper/2610.13945/) (2026) ## Abstract (arXiv) Training vision-language-action (VLA) policies for humanoid loco-manipulation is constrained by the high cost and complexity of collecting human teleoperation demonstrations. VLA policies fine-tuned in simulators have so far transferred poorly to humanoid loco-manipulation tasks. We present LEGS (Loco-manipulation via Embodied Gaussian Splatting), a hybrid simulator that composites a mesh foreground (robot, objects, props) over a photorealistic 3D Gaussian Splatting (3DGS) background reconstructed from a handheld scene capture. LEGS uses a procedural motion-primitive generator to synthesize labeled demonstrations at scale without human teleoperation, and a two-stage color calibration that aligns renders to the deployment camera. On a Unitree G1 humanoid robot, across three pick-and-place tasks of increasing whole-body difficulty and three VLA backbones (psi0, π0.5, 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 can 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/.