InterDiff: Generating 3D Human-Object Interactions with Physics-Informed Diffusion

TL;DR

InterDiff introduces a physics-informed diffusion framework to generate realistic, long-term 3D human-object interactions, addressing the challenge of modeling dynamic, whole-body interactions with various objects.

Contribution

The paper presents a novel diffusion-based approach with physics-informed correction for anticipating complex 3D human-object interactions.

Findings

Produces realistic and vivid 3D HOI predictions

Capable of long-term interaction generation

Effective across multiple datasets

Abstract

This paper addresses a novel task of anticipating 3D human-object interactions (HOIs). Most existing research on HOI synthesis lacks comprehensive whole-body interactions with dynamic objects, e.g., often limited to manipulating small or static objects. Our task is significantly more challenging, as it requires modeling dynamic objects with various shapes, capturing whole-body motion, and ensuring physically valid interactions. To this end, we propose InterDiff, a framework comprising two key steps: (i) interaction diffusion, where we leverage a diffusion model to encode the distribution of future human-object interactions; (ii) interaction correction, where we introduce a physics-informed predictor to correct denoised HOIs in a diffusion step. Our key insight is to inject prior knowledge that the interactions under reference with respect to contact points follow a simple pattern and are easily predictable. Experiments on multiple human-object interaction datasets demonstrate the effectiveness of our method for this task, capable of producing realistic, vivid, and remarkably long-term 3D HOI predictions.