Immersive Social Interaction with VR and LLM-Assisted Humanoids - 2026-07-08 - Research Publication - The paper presents an immersive teleoperation framework that integrates voice-controlled locomotion, VR-based manipulation, and bidirectional social interaction for whole-body humanoid control. Results show that novice… - source: arxiv.org
A System for Fast, Resilient, and Adaptable Loco-Manipulation Behaviors on Humanoid Robots - 2026-06-24 - Research Publication - This thesis presents a robot-local, runtime-editable behavior authoring and runtime system. This behavior system has been deployed on many humanoid robots, such as Boston Dynamics' DRC Atlas, NASA's Valkyrie, IHMC and… - source: arxiv.org
Constraint-enhanced RL paper validates actuator-rate-aware control on Unitree H1 - 2026-05-05 - Research Publication - The paper proposes dynamic decoupled spherical radial squashing to handle heterogeneous actuator-rate constraints in reinforcement-learning policies. High-fidelity Isaac Lab simulations use Unitree H1 and G1 humanoid… - source: arxiv.org
Humanoid recovery paper trains balance-aware stand-up policy on Unitree H1 - 2026-03-09 - Research Publication - The paper embeds classical balance metrics such as capture point, center-of-mass state, and centroidal momentum into reinforcement-learning training. The recovery policy is trained on Unitree H1-2 in Isaac Lab and reports… - source: arxiv.org
Residual Control for Fast Recovery from Dynamics Shifts - 2026-03-08 - Research Publication - The authors propose a stability- aligned residual control architecture in which a reinforcement learning policy trained under nominal dynamics remains fixed at deployment, and adaptation occurs exclusively through a… - source: arxiv.org
EAGLE distillation paper validates unified whole-body control on Unitree H1 - 2026-02-03 - Research Publication - The EAGLE paper introduces embodiment-aware generalist-specialist distillation for unified humanoid whole-body control across robot embodiments. The framework is validated on robots including Unitree H1, Unitree G1, and… - source: arxiv.org
FRoM-W1 language-conditioned whole-body control paper includes Unitree H1 - 2026-01-19 - Research Publication - FRoM-W1 is an open-source framework for general humanoid whole-body motion control from natural-language instructions. The paper presents the framework across humanoid platforms including Unitree H1, Unitree G1, and… - source: arxiv.org
Supernumerary-limbs locomotion paper trains Unitree H1 humanoid gait - 2025-11-25 - Research Publication - The paper proposes a hierarchical controller for humanoid locomotion with supernumerary limbs. The low-level component trains a walking gait for a Unitree H1 humanoid using imitation learning and curriculum learning. - source: arxiv.org
Humanoid Everyday dataset includes Unitree H1 demonstrations for open-world humanoid manipulation - 2025-10-09 - Research Publication - The Humanoid Everyday repository describes a diverse dataset of Unitree G1 and H1 demonstrations recorded at 30 Hz across everyday humanoid tasks. It lists 260 scenarios covering loco-manipulation, tool use, deformables… - source: github.com
HansZ8/RoboJuDo - 2025-09-28 - Research Publication - RoboJuDo 🤖 A plug-and-play deploy framework for robots. Experiment configurations can be organized through config files. - source: github.com
RobotDancing paper validates long-horizon humanoid motion tracking on Unitree H1 - 2025-09-25 - Research Publication - RobotDancing trains residual-action reinforcement-learning policies for robust long-horizon humanoid motion tracking. The paper evaluates primarily on Unitree G1 and validates transfer on H1 and H1-2 hardware. - source: arxiv.org
RoMoCo toolbox paper simulates reduced-order locomotion controllers on Unitree H1 - 2025-09-23 - Research Publication - RoMoCo is an open-source C++ toolbox for reduced-order model-based planners and whole-body controllers for bipedal and humanoid robots. The paper demonstrates simulations on Cassie, Unitree H1, and G1 robots, with hardware… - source: arxiv.org
Latent Conditioned Loco-Manipulation Using Motion Priors - 2025-09-19 - Research Publication - This approach is inefficient for solving more complicated tasks where high-level goals, physical robot limitations and desired motion style might all need to be taken into account. APPLICATION TOHUMANOID ANDQUADRUPEDAL… - source: arxiv.org
JeffrinSam/JeffrinSam_G1 - 2025-07-22 - Research Publication - Task 1: Synthetic Data Generation for Unitree H1 - Objective : Generate synthetic joint position data for Unitree H1 to train a policy for pick-and-place tasks. JeffrinSam MTS Synthetic Data Generation – Generate synthetic… - source: github.com
ISRIndustrial/JeffrinSam_MTS - 2025-07-21 - Research Publication - Task 1: Synthetic Data Generation for Unitree H1 - Objective : Generate synthetic joint position data for Unitree H1 to train a policy for pick-and-place tasks. JeffrinSam MTS Synthetic Data Generation – Generate synthetic… - source: github.com
GenPO: Generative Diffusion Models Meet On-Policy Reinforcement Learning - 2025-05-24 - Research Publication - GenPO makes diffusion policies compatible with on-policy reinforcement learning through exact diffusion inversion and an invertible doubled-action mechanism. The method is tested on eight Isaac Lab benchmarks spanning… - source: arxiv.org
TD-GRPC locomotion paper trains diverse Unitree H1-2 humanoid behaviors - 2025-05-19 - Research Publication - TD-GRPC combines temporal-difference learning with a group relative policy constraint for humanoid locomotion. The paper shows Unitree H1-2 tasks including standing, sitting, walking, running, crawling, stair climbing… - source: arxiv.org
ALMI learns coordinated whole-body motion policies on Unitree H1 - 2025-04-19 - Research Publication - ALMI separates lower-body locomotion and upper-body motion imitation into adversarially trained policies for coordinated humanoid movement. The authors report robust locomotion and precise motion tracking in simulation and… - source: arxiv.org
Humanoid-COA paper evaluates zero-shot loco-manipulation on Unitree H1-2 - 2025-04-13 - Research Publication - Humanoid-COA combines multimodal foundation models with embodied chain-of-action reasoning for zero-shot humanoid loco-manipulation. Experiments use Unitree H1-2 and G1 humanoids in open-area and apartment environments. - source: arxiv.org
InternRobotics/HoST - 2025-04-03 - Research Publication - Code is coming soon. - \[2025-05\] High Torque Mini Pi is now supported by HoST! Special thanks to: legged gym and HIMLoco: The foundation for training and running codes. rsl rl: Reinforcement learning algorithm… - source: github.com
Dynamic-balance RL paper validates narrow-terrain locomotion on Unitree H1-2 - 2025-02-24 - Research Publication - The paper proposes a whole-body reinforcement-learning controller for narrow pathways, unexpected obstacles, and external disturbances using proprioception rather than external perception. The authors validate the method… - source: arxiv.org
A Unified and General Humanoid Whole-Body Controller for Versatile Locomotion - 2025-02-05 - Research Publication - Locomotion is a fundamental skill for humanoid robots. However, most existing works make locomotion a single, tedious, unextendable, and unconstrained movement. The work is explicitly tied to H1, making the product binding… - source: arxiv.org
Learning Humanoid Locomotion with Perceptive Internal Model - 2024-11-21 - Research Publication - In contrast to quadruped robots that can navigate diverse terrains using a blind policy, humanoid robots require accurate perception for stable locomotion due to their high degrees of freedom and inherently unstable… - source: arxiv.org
HOVER unifies multiple whole-body control modes for humanoid robots - 2024-10-28 - Research Publication - HOVER introduces a multi-mode policy distillation framework for navigation, loco-manipulation, and tabletop manipulation control modes. The paper uses full-body kinematic motion imitation as a shared abstraction so one… - source: arxiv.org