MVHOI: Bridge Multi-view Condition to Complex Human-Object Interaction Video Reenactment via 3D Foundation Model

TL;DR

This paper introduces MVHOI, a two-stage framework that leverages a 3D foundation model to enable realistic, complex human-object interaction video reenactment with multi-view references, surpassing prior methods in handling 3D manipulations.

Contribution

The paper proposes a novel two-stage HOI video reenactment framework that integrates multi-view conditions with a 3D foundation model for improved realism and complex object manipulation.

Findings

Outperforms prior approaches in complex 3D object manipulations

Generates long-duration HOI videos with high fidelity

Ensures appearance consistency across views

Abstract

Human-Object Interaction (HOI) video reenactment with realistic motion remains a frontier in expressive digital human creation. Existing approaches primarily handle simple image-plane motion (e.g., in-plane translations), struggling with complex non-planar manipulations like out-of-plane reorientation. In this paper, we propose MVHOI, a two-stage HOI video reenactment framework that bridges multi-view reference conditions and video foundation models via a 3D Foundation Model (3DFM). The 3DFM first produces view-consistent object priors conditioned on implicit motion dynamics across novel viewpoints. A controllable video generation model then synthesizes high-fidelity object texture by incorporating multi-view reference images, ensuring appearance consistency via a reasonable retrieval mechanism. By enabling these two stages to mutually reinforce one another during the inference phase, our framework shows superior performance in generating long-duration HOI videos with intricate object manipulations. Extensive experiments show substantial improvements over prior approaches, especially for HOI with complex 3D object manipulations.