NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models

TL;DR

NeuralField-LDM introduces a hierarchical latent diffusion model for efficient and high-quality 3D scene generation, enabling diverse applications like scene synthesis, inpainting, and style transfer.

Contribution

The paper presents a novel hierarchical diffusion framework that compresses 3D scene representations into latent spaces for improved generation quality and versatility.

Findings

Outperforms existing scene generation models in quality and efficiency

Enables diverse applications such as conditional generation and scene editing

Uses a neural field auto-encoder with latent diffusion for 3D scene synthesis

Abstract

Automatically generating high-quality real world 3D scenes is of enormous interest for applications such as virtual reality and robotics simulation. Towards this goal, we introduce NeuralField-LDM, a generative model capable of synthesizing complex 3D environments. We leverage Latent Diffusion Models that have been successfully utilized for efficient high-quality 2D content creation. We first train a scene auto-encoder to express a set of image and pose pairs as a neural field, represented as density and feature voxel grids that can be projected to produce novel views of the scene. To further compress this representation, we train a latent-autoencoder that maps the voxel grids to a set of latent representations. A hierarchical diffusion model is then fit to the latents to complete the scene generation pipeline. We achieve a substantial improvement over existing state-of-the-art scene generation models. Additionally, we show how NeuralField-LDM can be used for a variety of 3D content creation applications, including conditional scene generation, scene inpainting and scene style manipulation.