WINE: Wavelet-Guided GAN Inversion and Editing for High-Fidelity Refinement

TL;DR

WINE introduces a wavelet-guided approach to GAN inversion that effectively preserves high-frequency details, improving image reconstruction and editing quality by interpreting inversion in the frequency domain.

Contribution

It is the first to interpret GAN inversion in the frequency domain and employs wavelet coefficients for high-frequency information transfer.

Findings

WINE outperforms existing models in preserving high-frequency details.

WINE achieves a better balance between editability and reconstruction quality.

Experimental results demonstrate improved image quality and detail preservation.

Abstract

Recent advanced GAN inversion models aim to convey high-fidelity information from original images to generators through methods using generator tuning or high-dimensional feature learning. Despite these efforts, accurately reconstructing image-specific details remains as a challenge due to the inherent limitations both in terms of training and structural aspects, leading to a bias towards low-frequency information. In this paper, we look into the widely used pixel loss in GAN inversion, revealing its predominant focus on the reconstruction of low-frequency features. We then propose WINE, a Wavelet-guided GAN Inversion aNd Editing model, which transfers the high-frequency information through wavelet coefficients via newly proposed wavelet loss and wavelet fusion scheme. Notably, WINE is the first attempt to interpret GAN inversion in the frequency domain. Our experimental results showcase the precision of WINE in preserving high-frequency details and enhancing image quality. Even in editing scenarios, WINE outperforms existing state-of-the-art GAN inversion models with a fine balance between editability and reconstruction quality.