PA-HOI: A Physics-Aware Human and Object Interaction Dataset

2025-08-08

AI Summary

Existing HOI datasets emphasize functional affordances while neglecting how physical object properties—such as mass, shape, and size—influence long-term human motion dynamics. To address this gap, we introduce the first physics-driven human-object interaction (HOI) motion capture dataset, comprising 562 interaction sequences involving multi-gender subjects and 35 diverse physical objects. We systematically characterize the temporal influence of object physics on pose evolution, motion velocity, and interaction strategies—a novel contribution. Leveraging high-fidelity motion capture, parametric 3D object modeling, and biomechanical motion dynamics analysis, we explicitly embed physical constraints into a generative action framework to enable physics-aware transfer. Experiments demonstrate that our dataset significantly improves the physical plausibility and perceptual realism of synthesized motions. Moreover, it exhibits strong generalization capability in robotic manipulation planning and VR-based interactive applications.

Technology Category

• Humans and AI: Interaction Techniques and Devices
• Computer Vision: Motion & Tracking
• Intelligent Robots: Manipulation

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

The Human-Object Interaction (HOI) task explores the dynamic interactions between humans and objects in physical environments, providing essential biomechanical and cognitive-behavioral foundations for fields such as robotics, virtual reality, and human-computer interaction. However, existing HOI data sets focus on details of affordance, often neglecting the influence of physical properties of objects on human long-term motion. To bridge this gap, we introduce the PA-HOI Motion Capture dataset, which highlights the impact of objects' physical attributes on human motion dynamics, including human posture, moving velocity, and other motion characteristics. The dataset comprises 562 motion sequences of human-object interactions, with each sequence performed by subjects of different genders interacting with 35 3D objects that vary in size, shape, and weight. This dataset stands out by significantly extending the scope of existing ones for understanding how the physical attributes of different objects influence human posture, speed, motion scale, and interacting strategies. We further demonstrate the applicability of the PA-HOI dataset by integrating it with existing motion generation methods, validating its capacity to transfer realistic physical awareness.

Problem

• Explores how object physical properties affect human motion dynamics
• Addresses lack of physics-aware human-object interaction datasets
• Analyzes impact of object attributes on posture and movement

Innovation

• PA-HOI dataset captures physical attributes' impact
• Includes 562 sequences with varied 3D objects
• Enhances motion generation with physical awareness