3DTopia: Large Text-to-3D Generation Model with Hybrid Diffusion Priors

TL;DR

3DTopia introduces a fast, two-stage text-to-3D generation system that combines 3D and 2D diffusion priors to produce high-quality 3D assets within five minutes.

Contribution

It presents a novel hybrid diffusion prior approach with a two-stage process for efficient and detailed text-to-3D asset creation, leveraging a new cleaned and captioned large-scale dataset.

Findings

High-quality 3D assets generated within 5 minutes

Effective refinement of textures using 2D diffusion priors

Qualitative and quantitative evaluation demonstrates superior performance

Abstract

We present a two-stage text-to-3D generation system, namely 3DTopia, which generates high-quality general 3D assets within 5 minutes using hybrid diffusion priors. The first stage samples from a 3D diffusion prior directly learned from 3D data. Specifically, it is powered by a text-conditioned tri-plane latent diffusion model, which quickly generates coarse 3D samples for fast prototyping. The second stage utilizes 2D diffusion priors to further refine the texture of coarse 3D models from the first stage. The refinement consists of both latent and pixel space optimization for high-quality texture generation. To facilitate the training of the proposed system, we clean and caption the largest open-source 3D dataset, Objaverse, by combining the power of vision language models and large language models. Experiment results are reported qualitatively and quantitatively to show the performance of the proposed system. Our codes and models are available at https://github.com/3DTopia/3DTopia