SkillMimic: Learning Basketball Interaction Skills from Demonstrations

2024-08-12

AI Summary

Traditional HOI reinforcement learning relies on hand-crafted skill rewards, suffering from poor generalization and limited scalability. This paper proposes the first unified data-driven HOI imitation reward framework, eliminating explicit reward engineering and enabling end-to-end learning of diverse basketball interaction skills directly from demonstrations. Methodologically, we integrate behavioral cloning with hierarchical imitation learning, curating a 35-minute high-fidelity basketball motion dataset and employing a hierarchical control architecture to learn multiple skills within a single policy with seamless switching. Key contributions include: (1) the first HOI-aware imitation reward mechanism; (2) automatic improvement in skill diversity and cross-task generalization as dataset scale increases; and (3) successful acquisition of multi-style skills—including dribbling, layups, and shooting—and their composition into long-horizon complex tasks such as continuous scoring sequences.

Technology Category

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

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Traditional reinforcement learning methods for human-object interaction (HOI) rely on labor-intensive, manually designed skill rewards that do not generalize well across different interactions. We introduce SkillMimic, a unified data-driven framework that fundamentally changes how agents learn interaction skills by eliminating the need for skill-specific rewards. Our key insight is that a unified HOI imitation reward can effectively capture the essence of diverse interaction patterns from HOI datasets. This enables SkillMimic to learn a single policy that not only masters multiple interaction skills but also facilitates skill transitions, with both diversity and generalization improving as the HOI dataset grows. For evaluation, we collect and introduce two basketball datasets containing approximately 35 minutes of diverse basketball skills. Extensive experiments show that SkillMimic successfully masters a wide range of basketball skills including stylistic variations in dribbling, layup, and shooting. Moreover, these learned skills can be effectively composed by a high-level controller to accomplish complex and long-horizon tasks such as consecutive scoring, opening new possibilities for scalable and generalizable interaction skill learning. Project page: https://ingrid789.github.io/SkillMimic/

Problem

• Eliminates need for manual skill-specific rewards in HOI learning
• Learns diverse basketball skills from unified imitation rewards
• Enables complex task composition via high-level controller

Innovation

• Unified data-driven framework eliminates skill-specific rewards
• Single policy learns diverse basketball interaction skills
• High-level controller composes skills for complex tasks