Learning Athletic Humanoid Tennis Skills from Imperfect Human Motion Data

2026-03-13

AI Summary

This work proposes a method for learning high-dynamic tennis striking skills from fragmented human motion data in the absence of complete demonstrations. By refining and recombining incomplete human motion priors, and integrating them with reinforcement learning policy optimization and robust sim-to-real transfer techniques, the approach constructs a generalizable hitting policy. It is the first to effectively extract useful priors from non-complete, highly variable human motion sequences, enabling the Unitree G1 humanoid robot to achieve stable multi-shot rallies in real-world environments. The resulting policy exhibits both high dynamic responsiveness and naturalistic movement style, demonstrating significant progress toward human-like agility and coordination in complex, interactive tasks.

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
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Human athletes demonstrate versatile and highly-dynamic tennis skills to successfully conduct competitive rallies with a high-speed tennis ball. However, reproducing such behaviors on humanoid robots is difficult, partially due to the lack of perfect humanoid action data or human kinematic motion data in tennis scenarios as reference. In this work, we propose LATENT, a system that Learns Athletic humanoid TEnnis skills from imperfect human motioN daTa. The imperfect human motion data consist only of motion fragments that capture the primitive skills used when playing tennis rather than precise and complete human-tennis motion sequences from real-world tennis matches, thereby significantly reducing the difficulty of data collection. Our key insight is that, despite being imperfect, such quasi-realistic data still provide priors about human primitive skills in tennis scenarios. With further correction and composition, we learn a humanoid policy that can consistently strike incoming balls under a wide range of conditions and return them to target locations, while preserving natural motion styles. We also propose a series of designs for robust sim-to-real transfer and deploy our policy on the Unitree G1 humanoid robot. Our method achieves surprising results in the real world and can stably sustain multi-shot rallies with human players. Project page: https://zzk273.github.io/LATENT/

Problem

• humanoid robotics
• tennis skills
• imperfect motion data
• motion imitation
• athletic behavior

Innovation

• humanoid robotics
• imitation learning
• imperfect motion data
• sim-to-real transfer
• athletic skills