Inclusive Interactive Collisions for Multi-View Consistent Compositional 3D Generation

2026-06-23

AI Summary

Existing 3D generation methods struggle to model plausible interactions among multiple objects and often suffer from cross-view inconsistencies. This work proposes I2C-3D, a novel approach that introduces, for the first time, an inclusive interaction-collision strategy to guide Gaussian primitives toward natural interactions within physically plausible regions. Additionally, it designs a multi-view adaptive score distillation sampling mechanism that effectively distills the multi-view consistency priors of diffusion models by modulating attention maps between instance and spatial tokens. The proposed method significantly enhances both generation quality and view consistency, enabling high-fidelity, editable, and compositional 3D scene synthesis.

Abstract

Recent breakthroughs in 3D generation have advanced notably with the development of text-to-image diffusion model. However, existing methods remain two practical challenges: (1) They primarily generate single 3D object, but struggle to generate multi-object compositional 3D assets due to the lack of the modeling for Gaussian primitives in reasonable interactions. (2) They often suffer from cross-view inconsistency during 3D optimization, as Score Distillation Sampling inherently performs on each single view, inevitably resulting in cross-view hallucinations. To solve above issues, we propose I2C-3D, a novel optimization-based method to generate multi-view consistent compositional 3D assets with reasonable interactions. Specifically, we propose an Inclusive Interactive Collisions strategy to guide Gaussian primitives appearing in reasonable interaction regions naturally, thereby ensuring objects in the compositional scene interact in a physically plausible and visually coherent way. Additionally, to enhance multi-view consistency, Multi-View Adaptive Score Distillation Sampling is devised to distill multi-view consistency prior and layout prior from pre-trained diffusion model by modulating attention map of instance token and spatial token across viewpoints. Benefiting from above elaborate designs, I2C-3D not only generates high-fidelity multi-view consistent compositional 3D assets but also supports 3D editing flexibly, facilitating complex scene generation. Extensive experiments demonstrate our I2C-3D outperforms existing methods in generation quality and multi-view consistency.

Problem

• compositional 3D generation
• multi-object interaction
• cross-view inconsistency
• Gaussian primitives
• Score Distillation Sampling

Innovation

• Inclusive Interactive Collisions
• Multi-View Consistency
• Compositional 3D Generation
• Score Distillation Sampling
• Gaussian Primitives