Part-aware Shape Generation with Latent 3D Diffusion of Neural Voxel Fields

TL;DR

This paper introduces a latent 3D diffusion model for neural voxel fields that enables high-resolution, part-aware shape generation with detailed textures and geometry, outperforming current methods.

Contribution

The paper proposes a novel latent 3D diffusion process and a part-aware shape decoder for improved, high-resolution, part-aware neural voxel shape generation.

Findings

Outperforms state-of-the-art methods in part-aware shape generation

Enables high-resolution generation with detailed textures and geometry

Demonstrates superior generative capabilities across multiple data classes

Abstract

This paper presents a novel latent 3D diffusion model for the generation of neural voxel fields, aiming to achieve accurate part-aware structures. Compared to existing methods, there are two key designs to ensure high-quality and accurate part-aware generation. On one hand, we introduce a latent 3D diffusion process for neural voxel fields, enabling generation at significantly higher resolutions that can accurately capture rich textural and geometric details. On the other hand, a part-aware shape decoder is introduced to integrate the part codes into the neural voxel fields, guiding the accurate part decomposition and producing high-quality rendering results. Through extensive experimentation and comparisons with state-of-the-art methods, we evaluate our approach across four different classes of data. The results demonstrate the superior generative capabilities of our proposed method in part-aware shape generation, outperforming existing state-of-the-art methods.