Face Super-Resolution with Progressive Embedding of Multi-scale Face Priors

TL;DR

This paper introduces a recurrent convolutional network framework for face super-resolution that progressively incorporates global shape and local texture priors, leading to improved facial detail recovery.

Contribution

It presents a novel recurrent network that progressively embeds multi-scale facial priors, including landmarks and action units, for enhanced super-resolution performance.

Findings

Outperforms state-of-the-art methods in image quality

Better restoration of facial details and textures

AU classification as a new quantitative metric

Abstract

The face super-resolution (FSR) task is to reconstruct high-resolution face images from low-resolution inputs. Recent works have achieved success on this task by utilizing facial priors such as facial landmarks. Most existing methods pay more attention to global shape and structure information, but less to local texture information, which makes them cannot recover local details well. In this paper, we propose a novel recurrent convolutional network based framework for face super-resolution, which progressively introduces both global shape and local texture information. We take full advantage of the intermediate outputs of the recurrent network, and landmarks information and facial action units (AUs) information are extracted in the output of the first and second steps respectively, rather than low-resolution input. Moreover, we introduced AU classification results as a novel quantitative metric for facial details restoration. Extensive experiments show that our proposed method significantly outperforms state-of-the-art FSR methods in terms of image quality and facial details restoration.