Papers › A Mountain-Shaped Single-Stage Network for Accurate Image Restoration

A Mountain-Shaped Single-Stage Network for Accurate Image Restoration

9 May 2023arXiv:2305.05146archive 2025-07-28

Hu Gao, Jing Yang, Ying Zhang, Ning Wang, Jingfan Yang, Depeng Dang

Image restoration is the task of aiming to obtain a high-quality image from a corrupt input image, such as deblurring and deraining. In image restoration, it is typically necessary to maintain a complex balance between spatial details and contextual information. Although a multi-stage network can optimally balance these competing goals and achieve significant performance, this also increases the system's complexity. In this paper, we propose a mountain-shaped single-stage design base on a simple U-Net architecture, which removes or replaces unnecessary nonlinear activation functions to achieve the above balance with low system complexity. Specifically, we propose a feature fusion middleware (FFM) mechanism as an information exchange component between the encoder-decoder architectural levels. It seamlessly integrates upper-layer information into the adjacent lower layer, sequentially down to the lowest layer. Finally, all information is fused into the original image resolution manipulation level. This preserves spatial details and integrates contextual information, ensuring high-quality image restoration. In addition, we propose a multi-head attention middle block (MHAMB) as a bridge between the encoder and decoder to capture more global information and surpass the limitations of the receptive field of CNNs. Extensive experiments demonstrate that our approach, named as M3SNet, outperforms previous state-of-the-art models while using less than half the computational costs, for several image restoration tasks, such as image deraining and deblurring.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

Tombs98/M3SNet officialpytorchNOASSERTION report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DeblurringDecoderImage DeblurringImage RestorationRain RemovalSingle Image Deraining

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Deblurring GoPro M3SNet PSNR 33.74 #15 of 55 Archive leaderboard report
Image Deblurring GoPro M3SNet SSIM 0.967 #15 of 55 Archive leaderboard report
Image Deblurring HIDE M3SNet PSNR 31.49 #4 of 5 Archive leaderboard report
Image Deblurring HIDE (trained on GOPRO) M3SNet PSNR 31.49 #2 of 3 Archive leaderboard report
Image Deblurring HIDE (trained on GOPRO) M3SNet SSIM 0.951 #2 of 3 Archive leaderboard report
Single Image Deraining Rain100H M3SNet PSNR 30.64 #9 of 19 Archive leaderboard report
Single Image Deraining Rain100H M3SNet SSIM 0.892 #9 of 19 Archive leaderboard report
Single Image Deraining Rain100L M3SNet PSNR 40.04 #2 of 19 Archive leaderboard report
Single Image Deraining Rain100L M3SNet SSIM 0.985 #2 of 19 Archive leaderboard report
Single Image Deraining Test100 M3SNet PSNR 31.29 #2 of 12 Archive leaderboard report
Single Image Deraining Test100 M3SNet SSIM 0.903 #2 of 12 Archive leaderboard report
Single Image Deraining Test1200 M3SNet PSNR 33.52 #3 of 14 Archive leaderboard report
Single Image Deraining Test1200 M3SNet SSIM 0.925 #3 of 14 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

BASEConcatenated Skip ConnectionConvolutionLinear LayerMax PoolingReLUSoftmaxU-Net

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections