Repurposing 2D Diffusion Models for 3D Shape Completion

TL;DR

This paper introduces a novel framework that leverages 2D diffusion models for 3D shape completion by using a new 2D representation called Shape Atlas, enabling high-quality 3D reconstructions from limited data.

Contribution

The paper proposes Shape Atlas, a compact 2D representation that aligns 3D geometry with 2D diffusion models, facilitating effective 3D shape completion from incomplete point clouds.

Findings

High-quality shape completions on PCN and ShapeNet-55 datasets

Effective utilization of pretrained 2D diffusion models for 3D tasks

Application in creating artist-designed 3D meshes

Abstract

We present a framework that adapts 2D diffusion models for 3D shape completion from incomplete point clouds. While text-to-image diffusion models have achieved remarkable success with abundant 2D data, 3D diffusion models lag due to the scarcity of high-quality 3D datasets and a persistent modality gap between 3D inputs and 2D latent spaces. To overcome these limitations, we introduce the Shape Atlas, a compact 2D representation of 3D geometry that (1) enables full utilization of the generative power of pretrained 2D diffusion models, and (2) aligns the modalities between the conditional input and output spaces, allowing more effective conditioning. This unified 2D formulation facilitates learning from limited 3D data and produces high-quality, detail-preserving shape completions. We validate the effectiveness of our results on the PCN and ShapeNet-55 datasets. Additionally, we show the downstream application of creating artist-created meshes from our completed point clouds, further demonstrating the practicality of our method.