SelFSR: Self-Conditioned Face Super-Resolution in the Wild via Flow Field Degradation Network

TL;DR

SelFSR introduces a domain-adaptive degradation network with flow field warping and a self-conditioned super-resolution block, significantly improving real-world face super-resolution by better handling domain gaps and preserving identity.

Contribution

The paper proposes a novel flow field degradation network and self-conditioned super-resolution block, enhancing face super-resolution in the wild without relying on explicit facial priors.

Findings

Achieves state-of-the-art results on CelebA and real-world datasets.

Effectively preserves identity and perceptual quality in real-world images.

Outperforms existing methods in handling domain gaps.

Abstract

In spite of the success on benchmark datasets, most advanced face super-resolution models perform poorly in real scenarios since the remarkable domain gap between the real images and the synthesized training pairs. To tackle this problem, we propose a novel domain-adaptive degradation network for face super-resolution in the wild. This degradation network predicts a flow field along with an intermediate low resolution image. Then, the degraded counterpart is generated by warping the intermediate image. With the preference of capturing motion blur, such a model performs better at preserving identity consistency between the original images and the degraded. We further present the self-conditioned block for super-resolution network. This block takes the input image as a condition term to effectively utilize facial structure information, eliminating the reliance on explicit priors, e.g. facial landmarks or boundary. Our model achieves state-of-the-art performance on both CelebA and real-world face dataset. The former demonstrates the powerful generative ability of our proposed architecture while the latter shows great identity consistency and perceptual quality in real-world images.