HOI-Diff: Text-Driven Synthesis of 3D Human-Object Interactions using Diffusion Models

TL;DR

This paper introduces HOI-Diff, a modular diffusion model framework that synthesizes realistic 3D human-object interactions from text prompts, incorporating motion generation, contact prediction, and contact-aware guidance.

Contribution

The paper presents a novel modular diffusion-based approach with a dual-branch motion model and an independent affordance prediction model for text-driven 3D HOI synthesis.

Findings

Produces realistic 3D HOIs with diverse interactions

Achieves accurate contact between humans and objects

Demonstrates effectiveness on BEHAVE and OMOMO datasets

Abstract

We address the problem of generating realistic 3D human-object interactions (HOIs) driven by textual prompts. To this end, we take a modular design and decompose the complex task into simpler sub-tasks. We first develop a dual-branch diffusion model (HOI-DM) to generate both human and object motions conditioned on the input text, and encourage coherent motions by a cross-attention communication module between the human and object motion generation branches. We also develop an affordance prediction diffusion model (APDM) to predict the contacting area between the human and object during the interactions driven by the textual prompt. The APDM is independent of the results by the HOI-DM and thus can correct potential errors by the latter. Moreover, it stochastically generates the contacting points to diversify the generated motions. Finally, we incorporate the estimated contacting points into the classifier-guidance to achieve accurate and close contact between humans and objects. To train and evaluate our approach, we annotate BEHAVE dataset with text descriptions. Experimental results on BEHAVE and OMOMO demonstrate that our approach produces realistic HOIs with various interactions and different types of objects.