Image2StyleGAN++: How to Edit the Embedded Images?

TL;DR

Image2StyleGAN++ enhances image editing by combining noise optimization, local embedding, and activation manipulation, enabling high-quality global and local edits such as inpainting, style transfer, and attribute transfer.

Contribution

It introduces noise optimization and local embedding extensions to the existing Image2StyleGAN framework for improved image reconstruction and editing capabilities.

Findings

Significant PSNR improvement from 20 dB to 45 dB with noise optimization.

Enables high-quality local and global image edits.

Supports diverse applications like inpainting, style transfer, and attribute transfer.

Abstract

We propose Image2StyleGAN++, a flexible image editing framework with many applications. Our framework extends the recent Image2StyleGAN in three ways. First, we introduce noise optimization as a complement to the $W^+$ latent space embedding. Our noise optimization can restore high-frequency features in images and thus significantly improves the quality of reconstructed images, e.g. a big increase of PSNR from 20 dB to 45 dB. Second, we extend the global $W^+$ latent space embedding to enable local embeddings. Third, we combine embedding with activation tensor manipulation to perform high-quality local edits along with global semantic edits on images. Such edits motivate various high-quality image editing applications, e.g. image reconstruction, image inpainting, image crossover, local style transfer, image editing using scribbles, and attribute level feature transfer. Examples of the edited images are shown across the paper for visual inspection.