Fourier-Guided Attention Upsampling for Image Super-Resolution

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Choi, Daejune, No, Youchan, Lee, Jinhyung, Kim, Duksu
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916913166155776
author Choi, Daejune
No, Youchan
Lee, Jinhyung
Kim, Duksu
author_facet Choi, Daejune
No, Youchan
Lee, Jinhyung
Kim, Duksu
contents We propose Frequency-Guided Attention (FGA), a lightweight upsampling module for single image super-resolution. Conventional upsamplers, such as Sub-Pixel Convolution, are efficient but frequently fail to reconstruct high-frequency details and introduce aliasing artifacts. FGA addresses these issues by integrating (1) a Fourier feature-based Multi-Layer Perceptron (MLP) for positional frequency encoding, (2) a cross-resolution Correlation Attention Layer for adaptive spatial alignment, and (3) a frequency-domain L1 loss for spectral fidelity supervision. Adding merely 0.3M parameters, FGA consistently enhances performance across five diverse super-resolution backbones in both lightweight and full-capacity scenarios. Experimental results demonstrate average PSNR gains of 0.12~0.14 dB and improved frequency-domain consistency by up to 29%, particularly evident on texture-rich datasets. Visual and spectral evaluations confirm FGA's effectiveness in reducing aliasing and preserving fine details, establishing it as a practical, scalable alternative to traditional upsampling methods.
format Preprint
id arxiv_https___arxiv_org_abs_2508_10616
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fourier-Guided Attention Upsampling for Image Super-Resolution
Choi, Daejune
No, Youchan
Lee, Jinhyung
Kim, Duksu
Computer Vision and Pattern Recognition
Artificial Intelligence
We propose Frequency-Guided Attention (FGA), a lightweight upsampling module for single image super-resolution. Conventional upsamplers, such as Sub-Pixel Convolution, are efficient but frequently fail to reconstruct high-frequency details and introduce aliasing artifacts. FGA addresses these issues by integrating (1) a Fourier feature-based Multi-Layer Perceptron (MLP) for positional frequency encoding, (2) a cross-resolution Correlation Attention Layer for adaptive spatial alignment, and (3) a frequency-domain L1 loss for spectral fidelity supervision. Adding merely 0.3M parameters, FGA consistently enhances performance across five diverse super-resolution backbones in both lightweight and full-capacity scenarios. Experimental results demonstrate average PSNR gains of 0.12~0.14 dB and improved frequency-domain consistency by up to 29%, particularly evident on texture-rich datasets. Visual and spectral evaluations confirm FGA's effectiveness in reducing aliasing and preserving fine details, establishing it as a practical, scalable alternative to traditional upsampling methods.
title Fourier-Guided Attention Upsampling for Image Super-Resolution
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2508.10616