AKDT: Adaptive Kernel Dilation Transformer for Effective Image Denoising

Open paper
Benchmark Model Rank Results
color-image-denoising-on-cbsd68-sigma15AKDT#1PSNR: 34.64
color-image-denoising-on-cbsd68-sigma25AKDT#1PSNR: 31.94
color-image-denoising-on-cbsd68-sigma50AKDT#5PSNR: 28.68
color-image-denoising-on-mcmaster-sigma15AKDT#1PSNR: 36.71
color-image-denoising-on-mcmaster-sigma25AKDT#1PSNR: 34.21
color-image-denoising-on-mcmaster-sigma50AKDT#1PSNR: 30.95
color-image-denoising-on-urban100-sigma15-1AKDT#1Average PSNR: 35.63PSNR: 35.63
color-image-denoising-on-urban100-sigma25AKDT#1PSNR: 33.14
color-image-denoising-on-urban100-sigma50AKDT#6PSNR: 29.82
image-denoising-on-siddAKDT#13PSNR (sRGB): 39.70SSIM (sRGB): 0.961
image-denoising-on-urban100-sigma15AKDT#2Average PSNR: 35.64
image-denoising-on-urban100-sigma50AKDT#4PSNR: 29.82

Rank counts only results with a code link.