WildFusion: Learning 3D-Aware Latent Diffusion Models in View Space

TL;DR

WildFusion introduces a novel 3D-aware image synthesis method using latent diffusion models in view space, eliminating the need for posed images or canonical representations, and achieves high-quality, 3D-consistent results from in-the-wild data.

Contribution

The paper proposes WildFusion, a 3D-aware latent diffusion model trained without multiview supervision, enabling scalable 3D content creation from unposed, in-the-wild images.

Findings

Outperforms recent GAN-based methods in 3D consistency and quality.

Learns 3D representations without multiview or pose supervision.

Enables novel view synthesis from unstructured image datasets.

Abstract

Modern learning-based approaches to 3D-aware image synthesis achieve high photorealism and 3D-consistent viewpoint changes for the generated images. Existing approaches represent instances in a shared canonical space. However, for in-the-wild datasets a shared canonical system can be difficult to define or might not even exist. In this work, we instead model instances in view space, alleviating the need for posed images and learned camera distributions. We find that in this setting, existing GAN-based methods are prone to generating flat geometry and struggle with distribution coverage. We hence propose WildFusion, a new approach to 3D-aware image synthesis based on latent diffusion models (LDMs). We first train an autoencoder that infers a compressed latent representation, which additionally captures the images' underlying 3D structure and enables not only reconstruction but also novel view synthesis. To learn a faithful 3D representation, we leverage cues from monocular depth prediction. Then, we train a diffusion model in the 3D-aware latent space, thereby enabling synthesis of high-quality 3D-consistent image samples, outperforming recent state-of-the-art GAN-based methods. Importantly, our 3D-aware LDM is trained without any direct supervision from multiview images or 3D geometry and does not require posed images or learned pose or camera distributions. It directly learns a 3D representation without relying on canonical camera coordinates. This opens up promising research avenues for scalable 3D-aware image synthesis and 3D content creation from in-the-wild image data. See https://katjaschwarz.github.io/wildfusion for videos of our 3D results.