StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation

TL;DR

StyleSDF is a novel method that combines SDF-based 3D modeling with style-based 2D generation to produce high-resolution, view-consistent images and detailed 3D shapes from single-view RGB data.

Contribution

It introduces a new approach that merges SDF-based 3D representation with StyleGAN2 for high-resolution, 3D-consistent image and shape generation from limited data.

Findings

Achieves 1024x1024 view-consistent images.

Produces detailed 3D surfaces with high geometric quality.

Outperforms state-of-the-art in visual and geometric metrics.

Abstract

We introduce a high resolution, 3D-consistent image and shape generation technique which we call StyleSDF. Our method is trained on single-view RGB data only, and stands on the shoulders of StyleGAN2 for image generation, while solving two main challenges in 3D-aware GANs: 1) high-resolution, view-consistent generation of the RGB images, and 2) detailed 3D shape. We achieve this by merging a SDF-based 3D representation with a style-based 2D generator. Our 3D implicit network renders low-resolution feature maps, from which the style-based network generates view-consistent, 1024x1024 images. Notably, our SDF-based 3D modeling defines detailed 3D surfaces, leading to consistent volume rendering. Our method shows higher quality results compared to state of the art in terms of visual and geometric quality.