DFE-Net: A Dual-Frequency Enhancement Network for Low-Light and Overexposed Image Restoration
Electronics · 2026-06-01 · 15 (11) · 2398
DOI: 10.3390/electronics15112398 ↗At a glance
Journal article in Electronics, published online on 2026-06-01. Written by Shengyou Zhou and 5 co-authors. 43 references. The full text is freely available (open access, CC BY 4.0). Funded by: National Natural Science Foundation of China, China Postdoctoral Science Foundation, Science and Technology Research Program of Chongqing Municipal Education Commission.
Abstract (original language)
In practical imaging applications, low-light and overexposure are two common types of image degradation problems with inherent conflicts, and existing methods struggle to achieve accurate restoration of both degradations within a unified framework. To address this challenge, this paper proposes DFE-Net based on explicit frequency decoupling. The network adopts a symmetric U-Net architecture and embeds discrete wavelet transform (DWT) and inverse discrete wavelet transform (IWT) to construct an explicit dual-frequency processing mechanism, which optimizes the low-frequency information carrying global illumination and the high-frequency information containing detailed textures, respectively. In the encoder, DWT decouples features into low-frequency and high-frequency sub-bands and feeds them into dedicated enhancement modules. The low-frequency enhancement block integrates SS2D and a gated convolutional feed-forward network to efficiently model global contextual dependencies with linear complexity and accurately restore image illumination and contrast; the high-frequency enhancement block adopts CMT attention combined with a matching convolutional feed-forward network, enabling the detail restoration process to be guided by the optimized low-frequency information and ensuring the collaborative optimization of global structure and local textures. The decoder completes the reconstruction and fusion of the processed sub-bands through IWT. The quantitative and qualitative experimental results on the MSEC, SICE, and LOLv1 datasets demonstrate that DFE-Net achieves or surpasses existing state-of-the-art methods in various metrics while maintaining low model complexity.
Funding
- National Natural Science Foundation of China
- China Postdoctoral Science Foundation
- Science and Technology Research Program of Chongqing Municipal Education Commission
- PLA Joint Logistics Support Force University of Engineering Youth Independent Innovation Fund Project
Cite
APA
Zhou, S., Chen, H., Cui, W., Chen, S., Wu, Z., & Chen, Y. (2026). DFE-Net: A Dual-Frequency Enhancement Network for Low-Light and Overexposed Image Restoration. Electronics, 15(11), 2398. https://doi.org/10.3390/electronics15112398
Vancouver
Zhou S, Chen H, Cui W, Chen S, Wu Z, Chen Y. DFE-Net: A Dual-Frequency Enhancement Network for Low-Light and Overexposed Image Restoration. Electronics. 2026;15(11):2398. doi:10.3390/electronics15112398
BibTeX
@article{Zhou2026,
author = {Zhou, Shengyou and Chen, Han and Cui, Wen and Chen, Shiming and Wu, Zhaojie and Chen, Yan},
title = {DFE-Net: A Dual-Frequency Enhancement Network for Low-Light and Overexposed Image Restoration},
journal = {Electronics},
year = {2026},
volume = {15},
number = {11},
pages = {2398},
doi = {10.3390/electronics15112398}
}
EUVIMED