EGM: Efficiently Learning General Motion Tracking Policy for High Dynamic Humanoid Whole-Body Control

• URL: http://arxiv.org/abs/2512.19043v1
• Date: Mon, 22 Dec 2025 05:25:24 GMT
• Title: EGM: Efficiently Learning General Motion Tracking Policy for High Dynamic Humanoid Whole-Body Control
• Abstract summary: EGM is a framework that enables efficient learning of a general motion tracking policy.<n>We develop a three-stage curriculum training flow to progressively enhance the policy's robustness against disturbances.<n>Despite training on only 4.08 hours of data, EGM generalized robustly across 49.25 hours of test motions, outperforming baselines on both routine and highly dynamic tasks.
• Score: 26.475304854903516
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Learning a general motion tracking policy from human motions shows great potential for versatile humanoid whole-body control. Conventional approaches are not only inefficient in data utilization and training processes but also exhibit limited performance when tracking highly dynamic motions. To address these challenges, we propose EGM, a framework that enables efficient learning of a general motion tracking policy. EGM integrates four core designs. Firstly, we introduce a Bin-based Cross-motion Curriculum Adaptive Sampling strategy to dynamically orchestrate the sampling probabilities based on tracking error of each motion bin, eficiently balancing the training process across motions with varying dificulty and durations. The sampled data is then processed by our proposed Composite Decoupled Mixture-of-Experts (CDMoE) architecture, which efficiently enhances the ability to track motions from different distributions by grouping experts separately for upper and lower body and decoupling orthogonal experts from shared experts to separately handle dedicated features and general features. Central to our approach is a key insight we identified: for training a general motion tracking policy, data quality and diversity are paramount. Building on these designs, we develop a three-stage curriculum training flow to progressively enhance the policy's robustness against disturbances. Despite training on only 4.08 hours of data, EGM generalized robustly across 49.25 hours of test motions, outperforming baselines on both routine and highly dynamic tasks.

Related papers

• Extreme-RGMT: Continual Learning of Highly Dynamic Skills for Robust Generalist Humanoid Control [12.247654049983522]
  We introduce Extreme-RGMT, a two-stage continual learning framework for robust generalist humanoid control.<n>Extreme-RGMT learns a generalist motion-tracking base policy from diverse multi-source motion data, then employs an asymmetric skill acquisition and capability consolidation mechanism.<n>Experiments show that Extreme-RGMT achieves state-of-the-art generalist whole-body motion-tracking performance.
  arXiv  Detail & Related papers  (2026-07-22T13:10:22Z)
• What Matters in Humanoid General Motion Tracking? An Empirical Study [2.728639503114606]
  We develop YAHMP, an open-source framework for training, evaluating, and deploying whole-body motion tracking policies on the Unitree G1.<n>We evaluate the resulting policies on a test set of retargeted human motions and compare the nominal policy with TWIST2 as an external baseline trained on the same motion set.
  arXiv  Detail & Related papers  (2026-07-22T08:37:19Z)
• SPARD: Self-Paced Curriculum for RL Alignment via Integrating Reward Dynamics and Data Utility [71.76390626651254]
  We propose SPARD, a framework that establishes an automated, self-paced curriculum by perceiving learning progress to adjust dynamically multi-objective reward weights and data importance.<n>Extensive experiments across multiple benchmarks demonstrate that SPARD significantly enhances model capabilities across all domains.
  arXiv  Detail & Related papers  (2026-04-09T05:37:22Z)
• Towards Generalizable Robotic Manipulation in Dynamic Environments [48.77270850350943]
  Vision-Language-Action (VLA) models excel in static manipulation but struggle in dynamic environments with moving trajectories.<n>We introduce DOMINO, a large-scale dataset and benchmark for general dynamic manipulation.<n>We also propose PUMA, a dynamics-aware VLA architecture.
  arXiv  Detail & Related papers  (2026-03-16T17:59:57Z)
• UniMotion: A Unified Motion Framework for Simulation, Prediction and Planning [60.55873455475112]
  We propose UniMotion, a unified motion framework that captures shared structures across motion tasks.<n>UniMotion employs dedicated interaction modes and tailored training strategies to simultaneously support these motion tasks.<n>With further fine-tuning, UniMotion achieves state-of-the-art performance across a range of motion tasks, establishing it as a versatile and scalable solution for autonomous driving.
  arXiv  Detail & Related papers  (2026-01-31T07:12:26Z)
• One-shot Humanoid Whole-body Motion Learning [18.375746497945023]
  Methods typically require multiple training samples per motion category.<n>We propose a novel approach that trains effective humanoid motion policies using only a single non-walking target motion sample.
  arXiv  Detail & Related papers  (2025-10-29T07:48:10Z)
• Curriculum Imitation Learning of Distributed Multi-Robot Policies [5.376095069606724]
  Learning control policies for multi-robot systems are a major challenge due to long-term coordination and the difficulty of obtaining realistic training data.<n>We propose a curriculum strategy that gradually increases the length of expert trajectories during training.<n>We also introduce a method to approximate the egocentric perception of each robot using only third-person global state demonstrations.
  arXiv  Detail & Related papers  (2025-09-29T17:31:48Z)
• UserRL: Training Interactive User-Centric Agent via Reinforcement Learning [104.63494870852894]
  Reinforcement learning (RL) has shown promise in training agentic models that engage in dynamic, multi-turn interactions.<n>We propose UserRL, a unified framework for training and evaluating user-centric abilities through standardized gym environments.
  arXiv  Detail & Related papers  (2025-09-24T03:33:20Z)
• Is Diversity All You Need for Scalable Robotic Manipulation? [50.747150672933316]
  We investigate the nuanced role of data diversity in robot learning by examining three critical dimensions-task (what to do), embodiment (which robot to use), and expert (who demonstrates)-challenging the conventional intuition of "more diverse is better"<n>We show that task diversity proves more critical than per-task demonstration quantity, benefiting transfer from diverse pre-training tasks to novel downstream scenarios.<n>We propose a distribution debiasing method to mitigate velocity ambiguity, the yielding GO-1-Pro achieves substantial performance gains of 15%, equivalent to using 2.5 times pre-training data.
  arXiv  Detail & Related papers  (2025-07-08T17:52:44Z)
• QuadrupedGPT: Towards a Versatile Quadruped Agent in Open-ended Worlds [51.05639500325598]
  We introduce QuadrupedGPT, designed to follow diverse commands with agility comparable to that of a pet.<n>Our agent shows proficiency in handling diverse tasks and intricate instructions, representing a significant step toward the development of versatile quadruped agents.
  arXiv  Detail & Related papers  (2024-06-24T12:14:24Z)
• LocoMuJoCo: A Comprehensive Imitation Learning Benchmark for Locomotion [20.545058017790428]
  Imitation Learning holds great promise for enabling agile locomotion in embodied agents. We present a novel benchmark designed to facilitate rigorous evaluation and comparison of IL algorithms. This benchmark encompasses a diverse set of environments, including quadrupeds, bipeds, and musculoskeletal human models.
  arXiv  Detail & Related papers  (2023-11-04T19:41:50Z)
• ALP: Action-Aware Embodied Learning for Perception [60.64801970249279]
  We introduce Action-Aware Embodied Learning for Perception (ALP) ALP incorporates action information into representation learning through a combination of optimizing a reinforcement learning policy and an inverse dynamics prediction objective. We show that ALP outperforms existing baselines in several downstream perception tasks.
  arXiv  Detail & Related papers  (2023-06-16T21:51:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.