Color Image Denoising on Kodak24 (test)
35.02PSNRDRANet
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DRANetsigma=152023.05 | 35.02 | — | — | — | |
| BRDNetsigma=152023.05 | 34.88 | — | — | — | |
| DudeNetsigma=152023.05 | 34.81 | — | — | — | |
| ADNetsigma=152023.05 | 34.76 | — | — | — | |
| AirNetsigma=152023.05 | 34.68 | — | — | — | |
| FFDNetNoise Level (sigma)=152017.10 | 34.63 | — | — | — | |
| FFDNetsigma=152023.05 | 34.63 | — | — | — | |
| DSNetBsigma=152023.05 | 34.63 | — | — | — | |
| IRCNNsigma=152023.05 | 34.56 | — | — | — | |
| CDnCNNNoise Level (sigma)=152017.10 | 34.48 | — | — | — | |
| CDnCNN-Ssigma=152023.05 | 34.48 | — | — | — | |
| BUIFDsigma=152023.05 | 34.41 | — | — | — | |
| CBM3DNoise Level (sigma)=152017.10 | 34.28 | — | — | — | |
| CBM3Dsigma=152023.05 | 34.28 | — | — | — | |
| DRANetsigma=252023.05 | 32.59 | — | — | — | |
| BRDNetsigma=252023.05 | 32.41 | — | — | — | |
| ADNetsigma=252023.05 | 32.26 | — | — | — | |
| DudeNetsigma=252023.05 | 32.26 | — | — | — | |
| AirNetsigma=252023.05 | 32.21 | — | — | — | |
| DSNetBsigma=252023.05 | 32.16 | — | — | — | |
| FFDNetNoise Level (sigma)=252017.10 | 32.13 | — | — | — | |
| FFDNetsigma=252023.05 | 32.13 | — | — | — | |
| CDnCNNNoise Level (sigma)=252017.10 | 32.03 | — | — | — | |
| CDnCNN-Ssigma=252023.05 | 32.03 | — | — | — | |
| IRCNNsigma=252023.05 | 32.03 | — | — | — | |
| RNANnoise level (sigma)=302020.01 | 31.86 | — | — | — | |
| SADNetnoise level (sigma)=302020.01 | 31.86 | — | — | — | |
| BUIFDsigma=252023.05 | 31.77 | — | — | — | |
| CBM3DNoise Level (sigma)=252017.10 | 31.68 | — | — | — | |
| CBM3Dsigma=252023.05 | 31.68 | — | — | — | |
| RIDNetnoise level (sigma)=302020.01 | 31.64 | — | — | — | |
| DnCNNnoise level (sigma)=302020.01 | 31.39 | — | — | — | |
| FFDNetnoise level (sigma)=302020.01 | 31.39 | — | — | — | |
| CBM3Dnoise level (sigma)=302020.01 | 30.89 | — | — | — | |
| FFDNetNoise Level (sigma)=352017.10 | 30.57 | — | — | — | |
| CDnCNNNoise Level (sigma)=352017.10 | 30.46 | — | — | — | |
| CBM3DNoise Level (sigma)=352017.10 | 29.9 | — | — | — | |
| MemNetnoise level (sigma)=302020.01 | 29.67 | — | — | — | |
| SADNetnoise level (sigma)=502020.01 | 29.64 | — | — | — | |
| RNANnoise level (sigma)=502020.01 | 29.58 | — | — | — | |
| DRANetsigma=502023.05 | 29.5 | — | — | — | |
| MC-WNNMNoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 29.31 | — | — | — | |
| RIDNetnoise level (sigma)=502020.01 | 29.25 | — | — | — | |
| BRDNetsigma=502023.05 | 29.22 | — | — | — | |
| DnCNNnoise level (sigma)=502020.01 | 29.16 | — | — | — | |
| FFDNetnoise level (sigma)=502020.01 | 29.1 | — | — | — | |
| ADNetsigma=502023.05 | 29.1 | — | — | — | |
| DudeNetsigma=502023.05 | 29.1 | — | — | — | |
| AirNetsigma=502023.05 | 29.06 | — | — | — | |
| DSNetBsigma=502023.05 | 29.05 | — | — | — | |
| FFDNetNoise Level (sigma)=502017.10 | 28.98 | — | — | — | |
| FFDNetsigma=502023.05 | 28.98 | — | — | — | |
| CDnCNNNoise Level (sigma)=502017.10 | 28.85 | — | — | — | |
| CDnCNN-Ssigma=502023.05 | 28.85 | — | — | — | |
| WNNM-1Noise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 28.84 | — | — | — | |
| WNNM-2Noise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 28.83 | — | — | — | |
| IRCNNsigma=502023.05 | 28.81 | — | — | — | |
| TNRDNoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 28.68 | — | — | — | |
| CBM3Dnoise level (sigma)=502020.01 | 28.63 | — | — | — | |
| MLPNoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 28.54 | — | — | — | |
| CBM3DNoise Level (sigma)=502017.10 | 28.46 | — | — | — | |
| CBM3Dsigma=502023.05 | 28.46 | — | — | — | |
| SADNetnoise level (sigma)=702020.01 | 28.28 | — | — | — | |
| WNNM-3Noise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 28.22 | — | — | — | |
| RNANnoise level (sigma)=702020.01 | 28.16 | — | — | — | |
| RIDNetnoise level (sigma)=702020.01 | 27.94 | — | — | — | |
| BUIFDsigma=502023.05 | 27.74 | — | — | — | |
| FFDNetnoise level (sigma)=702020.01 | 27.68 | — | — | — | |
| MemNetnoise level (sigma)=502020.01 | 27.65 | — | — | — | |
| DnCNNnoise level (sigma)=702020.01 | 27.64 | — | — | — | |
| FFDNetNoise Level (sigma)=752017.10 | 27.27 | — | — | — | |
| CBM3Dnoise level (sigma)=702020.01 | 27.27 | — | — | — | |
| CBM3DNoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 27.13 | — | — | — | |
| CBM3DNoise Level (sigma)=752017.10 | 26.82 | — | — | — | |
| MemNetnoise level (sigma)=702020.01 | 26.4 | — | — | — | |
| NCNoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 26.19 | — | — | — | |
| NINoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 25.24 | — | — | — | |
| CDnCNNNoise Level (sigma)=752017.10 | 25.04 | — | — | — | |
| DnCNNNoise level (sigma_r, sigma_g, sigma_b)=40, 20, 302017.05 | 20.58 | — | — | — | |
| ADFNet2025.12 | — | 34.77 | 32.22 | 29.06 | |
| AirNet2025.12 | — | 34.81 | 32.44 | 29.1 | |
| ART*MACs (G)=4220, Inf. (ms)=OOM, Image Size=512 x 512 x 32024.01 | — | 35.39 | 32.95 | 29.87 | |
| BRDNetTraining strategy=separate models2021.11 | — | 34.88 | 32.41 | 29.22 | |
| BRDNetTraining strategy=Noise-specific (separate models)2023.07 | — | 34.88 | 32.41 | 29.22 | |
| CBM3DTraining strategy=Noise-blind (various noise levels)2023.07 | — | 34.28 | 32.15 | 28.46 | |
| CGNetMACs (G)=444, Inf. (ms)=215, Image Size=512 x 512 x 32024.01 | — | 35.52 | 33.07 | 30.06 | |
| ClusIR2025.12 | — | 35.06 | 32.6 | 29.49 | |
| CODEMACs (G)=180, Inf. (ms)=600, Image Size=512 x 512 x 32024.01 | — | 35.32 | 32.88 | 29.82 | |
| CODETraining strategy=Noise-specific (separate models)2023.07 | — | 35.32 | 32.88 | 29.82 | |
| DMID-dTraining strategy=Noise-blind (various noise levels)2023.07 | — | 35.51 | 33.12 | 30.14 | |
| DnCNNTraining strategy=single model2021.11 | — | 34.6 | 32.14 | 28.95 | |
| DnCNNParams [M]=0.562023.03 | — | 34.6 | 32.14 | 28.95 | |
| DnCNNTraining strategy=Noise-blind (various noise levels)2023.07 | — | 34.6 | 32.14 | 28.95 | |
| DnCNNParams [M]=0.562024.11 | — | 34.6 | 32.14 | 28.95 | |
| DnCNN2025.12 | — | 34.6 | 32.14 | 28.95 | |
| DRUNetTraining strategy=single model2021.11 | — | 35.31 | 32.89 | 29.86 | |
| DRUNetParams [M]=32.642023.03 | — | 35.31 | 32.89 | 29.86 | |
| DRUNetTraining strategy=Noise-blind (various noise levels)2023.07 | — | 35.31 | 32.89 | 29.86 | |
| DRUNetParams [M]=32.642024.11 | — | 35.31 | 32.89 | 29.86 | |
| DSNetTraining strategy=single model2021.11 | — | 34.63 | 32.16 | 29.05 |