Papers › Exploring the potential of channel interactions for image restoration

Exploring the potential of channel interactions for image restoration

20 Dec 2023Knowledge-Based Systems 2023 12archive 2025-07-28

Yuning Cui, Alois Knoll

Image restoration aims to reconstruct a clear image from a degraded observation. Convolutional neural networks have achieved promising performance on this task. The usage of Transformer has recently made significant advancements in state-of-the-art performance by modeling long-range dependencies. However, these deep architectures primarily concentrate on enhancing representation learning for the spatial dimension, neglecting the significance of channel interactions. In this paper, we explore the potential of channel interactions for restoring images through our proposal of a dual-domain channel attention mechanism. To be specific, channel attention in the spatial domain allows each channel to amass valuable signals from adjacent channels under the guidance of learned dynamic weights. In order to effectively exploit the significant difference in infrequency between degraded and clean image pairs, we develop the implicit frequency domain channel attention to facilitate the integration of information from different frequencies. Extensive experiments demonstrate that the proposed network, dubbed ChaIR, achieves state-of-the-art performance on 13 benchmark datasets for five image restoration tasks, including image dehazing, image motion/defocus deblurring, image desnowing, and image deraining.

Paper PDF Code

Code

c-yn/ChaIR pytorch 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

Deblurring Image Deblurring Image Defocus Deblurring Image Dehazing Image Restoration Rain Removal Representation Learning Single Image Deraining Single Image Desnowing

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Image Deblurring	GoPro	ChaIR	PSNR	33.28	#25 of 55	Archive leaderboard	report
Image Deblurring	GoPro	ChaIR	SSIM	0.963	#25 of 55	Archive leaderboard	report
Image Dehazing	SOTS Indoor	ChaIR	PSNR	41.95	#5 of 34	Archive leaderboard	report
Image Dehazing	SOTS Indoor	ChaIR	SSIM	0.997	#5 of 34	Archive leaderboard	report
Image Dehazing	SOTS Outdoor	ChaIR	PSNR	40.73	#1 of 31	Archive leaderboard	report
Image Dehazing	SOTS Outdoor	ChaIR	SSIM	0.997	#1 of 31	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

Absolute Position Encodings Adam Attention BPE Dense Connections Dropout Label Smoothing Layer Normalization Linear Layer Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Transformer

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