Image Denoising on BSD200 (test)
36.36PSNRRC-Net
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RC-NetNoise level (σ)=102019.10 | 36.36 | 95.41 | 32.55 | |
| WIN5-RB-Ssigma=102017.07 | 35.83 | 94.94 | — | |
| WINNoise level (σ)=102019.10 | 35.83 | 94.94 | 28.72 | |
| WIN5-RB-Bsigma=102017.07 | 35.43 | 94.61 | — | |
| DnCNNsigma=102017.07 | 34.6 | 92.83 | — | |
| DnCNNNoise level (σ)=102019.10 | 34.6 | 92.83 | 13.49 | |
| WIN5-Rsigma=102017.07 | 34.43 | 92.43 | — | |
| WIN5sigma=102017.07 | 34.1 | 92.05 | — | |
| BM3Dsigma=102017.07 | 34.02 | 91.82 | — | |
| BM3DNoise level (σ)=102019.10 | 34.02 | 91.82 | 1.01 | |
| RED30noise level (sigma)=102016.03 | 33.63 | 93.19 | — | |
| WIN5-RB-Ssigma=302017.07 | 33.62 | 91.93 | 20.39 | |
| WINNoise level (σ)=302019.10 | 33.62 | 91.93 | 31.94 | |
| RED20noise level (sigma)=102016.03 | 33.59 | 93.1 | — | |
| RC-NetNoise level (σ)=302019.10 | 33.57 | 92.71 | 33.52 | |
| RED10noise level (sigma)=102016.03 | 33.49 | 92.9 | — | |
| PCLRnoise level (sigma)=102016.03 | 33.3 | 92.61 | — | |
| WIN5-RB-Bsigma=302017.07 | 33.27 | 92.63 | 15.82 | |
| WNNMnoise level (sigma)=102016.03 | 33.25 | 92.44 | — | |
| NCSRnoise level (sigma)=102016.03 | 33.09 | 92.26 | — | |
| PGPDnoise level (sigma)=102016.03 | 33.02 | 91.76 | — | |
| BM3Dnoise level (sigma)=102016.03 | 33.01 | 92.18 | — | |
| EPLLnoise level (sigma)=102016.03 | 33.01 | 92.55 | — | |
| RED-Netsigma=102017.07 | 32.96 | 89.63 | — | |
| RED-NetNoise level (σ)=102019.10 | 32.96 | 89.63 | 60.73 | |
| RC-NetNoise level (σ)=502019.10 | 32.48 | 91.12 | 33.52 | |
| WIN5-RB-Bsigma=502017.07 | 32.18 | 91.36 | 13.79 | |
| WIN5-RB-Ssigma=502017.07 | 31.79 | 88.31 | 21.79 | |
| WINNoise level (σ)=502019.10 | 31.79 | 88.31 | 23.77 | |
| RC-NetNoise level (σ)=702019.10 | 31.17 | 87.95 | 32.81 | |
| WIN5-RB-Bsigma=702017.07 | 31.07 | 89.62 | 13.17 | |
| WIN5-Rsigma=302017.07 | 30.94 | 86.44 | 15.78 | |
| WIN5-RB-Ssigma=702017.07 | 30.34 | 83.62 | 20.86 | |
| WINNoise level (σ)=702019.10 | 30.34 | 83.62 | 23.9 | |
| WIN5-Rsigma=502017.07 | 29.38 | 82.51 | 22.72 | |
| DnCNNsigma=302017.07 | 29.13 | 80.6 | 13.61 | |
| DnCNNNoise level (σ)=302019.10 | 29.13 | 80.6 | 13.48 | |
| RED-Netsigma=302017.07 | 29.05 | 80.49 | 69.25 | |
| RED-NetNoise level (σ)=302019.10 | 29.05 | 80.49 | 60.27 | |
| WIN5sigma=302017.07 | 28.93 | 79.87 | 15.36 | |
| BM3Dsigma=302017.07 | 28.57 | 78.23 | 1.67 | |
| WIN5sigma=502017.07 | 28.57 | 79.79 | 16.7 | |
| BM3DNoise level (σ)=302019.10 | 28.57 | 78.23 | 1.23 | |
| WIN5-Rsigma=702017.07 | 28.16 | 77.84 | 19.28 | |
| WIN5sigma=702017.07 | 27.98 | 78.75 | 16.1 | |
| RED30noise level (sigma)=302016.03 | 27.95 | 80.19 | — | |
| RED20noise level (sigma)=302016.03 | 27.9 | 79.93 | — | |
| RED10noise level (sigma)=302016.03 | 27.79 | 79.18 | — | |
| PCLRnoise level (sigma)=302016.03 | 27.54 | 78.27 | — | |
| WNNMnoise level (sigma)=302016.03 | 27.48 | 78.07 | — | |
| EPLLnoise level (sigma)=302016.03 | 27.38 | 78.25 | — | |
| PGPDnoise level (sigma)=302016.03 | 27.33 | 77.17 | — | |
| BM3Dnoise level (sigma)=302016.03 | 27.31 | 77.55 | — | |
| NCSRnoise level (sigma)=302016.03 | 27.23 | 77.38 | — | |
| DnCNNsigma=502017.07 | 26.99 | 72.89 | 13.76 | |
| DnCNNNoise level (σ)=502019.10 | 26.99 | 72.89 | 12.55 | |
| RED-Netsigma=502017.07 | 26.88 | 72.3 | 70.34 | |
| RED-NetNoise level (σ)=502019.10 | 26.88 | 72.3 | 66.62 | |
| RED-Netsigma=702017.07 | 26.66 | 71.08 | 69.99 | |
| RED-NetNoise level (σ)=702019.10 | 26.66 | 71.08 | 66.99 | |
| BM3Dsigma=502017.07 | 26.44 | 70.28 | 2.87 | |
| BM3DNoise level (σ)=502019.10 | 26.44 | 70.28 | 2.02 | |
| RED30noise level (sigma)=502016.03 | 25.75 | 71.67 | — | |
| RED20noise level (sigma)=502016.03 | 25.67 | 71.17 | — | |
| DnCNNsigma=702017.07 | 25.65 | 67.09 | 12.88 | |
| DnCNNNoise level (σ)=702019.10 | 25.65 | 67.09 | 13.42 | |
| RED10noise level (sigma)=502016.03 | 25.54 | 70.32 | — | |
| PCLRnoise level (sigma)=502016.03 | 25.3 | 69.47 | — | |
| WNNMnoise level (sigma)=502016.03 | 25.26 | 69.28 | — | |
| BM3Dsigma=702017.07 | 25.23 | 65.22 | 2.93 | |
| BM3DNoise level (σ)=702019.10 | 25.23 | 65.22 | 1.98 | |
| PGPDnoise level (sigma)=502016.03 | 25.18 | 68.41 | — | |
| EPLLnoise level (sigma)=502016.03 | 25.17 | 68.7 | — | |
| BM3Dnoise level (sigma)=502016.03 | 25.06 | 68.31 | — | |
| NCSRnoise level (sigma)=502016.03 | 24.95 | 67.77 | — | |
| RED30noise level (sigma)=702016.03 | 24.37 | 65.51 | — | |
| RED20noise level (sigma)=702016.03 | 24.33 | 65.21 | — | |
| RED10noise level (sigma)=702016.03 | 24.13 | 63.67 | — | |
| WNNMnoise level (sigma)=702016.03 | 23.95 | 63.46 | — | |
| PCLRnoise level (sigma)=702016.03 | 23.94 | 63.36 | — | |
| PGPDnoise level (sigma)=702016.03 | 23.89 | 62.45 | — | |
| BM3Dnoise level (sigma)=702016.03 | 23.82 | 62.4 | — | |
| EPLLnoise level (sigma)=702016.03 | 23.81 | 61.68 | — | |
| NCSRnoise level (sigma)=702016.03 | 23.58 | 61.66 | — |