Efficient Face Super-Resolution via Wavelet-based Feature Enhancement Network
Open Access
- 28 October 2024
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 4515-4523
Abstract
Face super-resolution aims to reconstruct a high-resolution face image from a low-resolution face image. Previous methods typically employ an encoder-decoder structure to extract facial structural features, where the direct downsampling inevitably introduces distortions, especially to high-frequency features such as edges. To address this issue, we propose a wavelet-based feature enhancement network, which mitigates feature distortion by losslessly decomposing the input feature into high and low-frequency components using the wavelet transform and processing them separately. To improve the efficiency of facial feature extraction, a full domain Transformer is further proposed to enhance local, regional, and global facial features. Such designs allow our method to perform better without stacking many modules as previous methods did. Experiments show that our method effectively balances performance, model size, and speed. Code link: https://github.com/PRIS-CV/WFEN.Keywords
All Related Versions
This publication has 23 references indexed in Scilit:
- Feature Modulation Transformer: Cross-Refinement of Global Representation via High-Frequency Prior for Image Super-Resolution2023
- Spatial-Frequency Mutual Learning for Face Super-Resolution2023
- Scaling Up Your Kernels to 31×31: Revisiting Large Kernel Design in CNNs2022
- Edge and identity preserving network for face super-resolutionNeurocomputing, 2021
- Wavelet Domain Generative Adversarial Network for Multi-scale Face HallucinationInternational Journal of Computer Vision, 2019
- The Unreasonable Effectiveness of Deep Features as a Perceptual Metric2018
- Deep Learning Face Attributes in the Wild2015
- SCface – surveillance cameras face databaseMultimedia Tools and Applications, 2009
- A universal image quality indexIEEE Signal Processing Letters, 2002
- A theory for multiresolution signal decomposition: the wavelet representationIEEE Transactions on Pattern Analysis and Machine Intelligence, 1989