Accurate Image Restoration with Attention Retractable Transformer

Open paper
Benchmark Model Rank Results
image-super-resolution-on-bsd100-2x-upscalingART–PSNR: 32.58SSIM: 0.9048
image-super-resolution-on-bsd100-3x-upscalingART–PSNR: 29.51SSIM: 0.8159
image-super-resolution-on-bsd100-4x-upscalingART–PSNR: 27.97SSIM: 0.7510
image-super-resolution-on-manga109-2x-upscalingART–PSNR: 40.24SSIM: 0.9808
image-super-resolution-on-manga109-3x-upscalingART–PSNR: 35.39SSIM: 0.9548
image-super-resolution-on-manga109-4x-upscalingART–SSIM: 0.9283PSNR: 32.31
image-super-resolution-on-set14-2x-upscalingART–PSNR: 34.59SSIM: 0.9267
image-super-resolution-on-set14-3x-upscalingART–PSNR: 31.02SSIM: 0.8541
image-super-resolution-on-set14-4x-upscalingART–PSNR: 29.16SSIM: 0.7958
image-super-resolution-on-set5-2x-upscalingART–PSNR: 38.56SSIM: 0.9629
image-super-resolution-on-set5-3x-upscalingART–PSNR: 35.07SSIM: 0.9325
image-super-resolution-on-set5-4x-upscalingART–PSNR: 33.04SSIM: 0.9051
image-super-resolution-on-urban100-2x-upscalingART–PSNR: 34.30SSIM: 0.9452
image-super-resolution-on-urban100-3x-upscalingART–PSNR: 30.10SSIM: 0.8871
image-super-resolution-on-urban100-4x-upscalingART–PSNR: 27.77SSIM: 0.8321

Rank counts only results with a code link.