Unleashing Guidance Without Classifiers for Human-Object Interaction Animation

2026-03-26

AI Summary

Generating realistic human-object interaction animations requires joint modeling of human dynamics and object geometry, yet existing approaches often rely on handcrafted contact priors. This work proposes LIGHT, a diffusion-forcing framework that leverages modality decomposition and asynchronous denoising scheduling to generate internal guidance signals directly from the denoising rhythm itself, enabling contact-aware synthesis without external classifiers. By exploiting the temporal structure of the denoising process, LIGHT enhances invariance to object shape and significantly outperforms current classifier-free guidance methods in terms of contact fidelity, interaction realism, and generalization to unseen objects and tasks.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Humans and AI: Interaction Techniques and Devices
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Generating realistic human-object interaction (HOI) animations remains challenging because it requires jointly modeling dynamic human actions and diverse object geometries. Prior diffusion-based approaches often rely on hand-crafted contact priors or human-imposed kinematic constraints to improve contact quality. We propose LIGHT, a data-driven alternative in which guidance emerges from the denoising pace itself, reducing dependence on manually designed priors. Building on diffusion forcing, we factor the representation into modality-specific components and assign individualized noise levels with asynchronous denoising schedules. In this paradigm, cleaner components guide noisier ones through cross-attention, yielding guidance without auxiliary classifiers. We find that this data-driven guidance is inherently contact-aware, and can be enhanced when training is augmented with a broad spectrum of synthetic object geometries, encouraging invariance of contact semantics to geometric diversity. Extensive experiments show that pace-induced guidance more effectively mirrors the benefits of contact priors than conventional classifier-free guidance, while achieving higher contact fidelity, more realistic HOI generation, and stronger generalization to unseen objects and tasks.

Problem

• human-object interaction
• animation generation
• contact modeling
• object geometry
• realistic animation

Innovation

• diffusion forcing
• asynchronous denoising
• classifier-free guidance
• contact-aware generation
• human-object interaction