A Kung Fu Athlete Bot That Can Do It All Day: Highly Dynamic, Balance-Challenging Motion Dataset and Autonomous Fall-Resilient Tracking

2026-02-14

AI Summary

Existing motion control systems for humanoid robots struggle to execute high-intensity, highly dynamic martial arts maneuvers and lack a unified framework for modeling non-steady states such as falls or enabling autonomous recovery. To address these limitations, this work introduces KungFuAthlete, a high-dynamic motion dataset derived from professional martial artists, whose joint velocities, linear velocities, and angular velocities substantially exceed those of existing benchmarks like LAFAN1, PHUMA, and AMASS. Furthermore, the study proposes a unified reinforcement learning policy that, for the first time, integrates high-dynamic motion tracking and autonomous fall recovery within a single framework. This approach enables synergistic optimization of agile execution and stable control, significantly enhancing the robustness and autonomy of humanoid robots in complex, dynamic environments.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Current humanoid motion tracking systems can execute routine and moderately dynamic behaviors, yet significant gaps remain near hardware performance limits and algorithmic robustness boundaries. Martial arts represent an extreme case of highly dynamic human motion, characterized by rapid center-of-mass shifts, complex coordination, and abrupt posture transitions. However, datasets tailored to such high-intensity scenarios remain scarce. To address this gap, we construct KungFuAthlete, a high-dynamic martial arts motion dataset derived from professional athletes'daily training videos. The dataset includes ground and jump subsets covering representative complex motion patterns. The jump subset exhibits substantially higher joint, linear, and angular velocities compared to commonly used datasets such as LAFAN1, PHUMA, and AMASS, indicating significantly increased motion intensity and complexity. Importantly, even professional athletes may fail during highly dynamic movements. Similarly, humanoid robots are prone to instability and falls under external disturbances or execution errors. Most prior work assumes motion execution remains within safe states and lacks a unified strategy for modeling unsafe states and enabling reliable autonomous recovery. We propose a novel training paradigm that enables a single policy to jointly learn high-dynamic motion tracking and fall recovery, unifying agile execution and stabilization within one framework. This framework expands robotic capability from pure motion tracking to recovery-enabled execution, promoting more robust and autonomous humanoid performance in real-world high-dynamic scenarios.

Problem

• humanoid motion tracking
• high-dynamic motion
• fall recovery
• motion dataset
• robotic robustness

Innovation

• high-dynamic motion dataset
• fall-resilient control
• unified policy learning
• humanoid motion tracking
• martial arts robotics