Image Denoising On Sidd

PSNR (sRGB)

SSIM (sRGB)

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	PSNR (sRGB)	SSIM (sRGB)
akdt-adaptive-kernel-dilation-transformer-for	39.70	0.961
learning-enriched-features-for-real-image	39.72	0.959
maxim-multi-axis-mlp-for-image-processing	39.96	0.960
hinet-half-instance-normalization-network-for	39.99	0.958
variational-denoising-network-toward-blind	39.28	0.956
real-image-denoising-with-feature-attention	38.71	0.951
dual-residual-attention-network-for-image	-	0.957
simple-baselines-for-image-restoration	40.30	0.961
dual-adversarial-network-toward-real-world	39.47	0.957
single-stage-adaptive-multi-attention-network	40.08	0.962
cascadedgaze-efficiency-in-global-context	40.39	0.964
transfer-learning-from-synthetic-to-real-2	38.95	0.952
nbnet-noise-basis-learning-for-image	-	0.973
restormer-efficient-transformer-for-high	40.02	0.960
toward-convolutional-blind-denoising-of-real	30.78	0.801
spatial-adaptive-network-for-single-image	39.46	0.957
selective-residual-m-net-for-real-image	39.72	0.959
kbnet-kernel-basis-network-for-image	40.35	0.972
cycleisp-real-image-restoration-via-improved	39.52	0.957
uformer-a-general-u-shaped-transformer-for	39.89	0.960
multi-stage-progressive-image-restoration	39.71	0.958
learning-to-generate-realistic-noisy-images-2	40.07	0.960