Image Denoising on DND benchmark (test)
39.77PSNRAINDNet(S)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AINDNet(S)Blind/Non-blind=Blind, Training Env.=Synthetic, Geometric Self-Ensemble=true2020.02 | 39.77 | 0.959 | |
| DANetLearning Paradigm=Supervised2023.03 | 39.58 | 0.955 | |
| InvDNLearning Paradigm=Supervised2023.03 | 39.57 | 0.952 | |
| AINDNet(S)Blind/Non-blind=Blind, Training Env.=Synthetic, Geometric Self-Ensemble=false2020.02 | 39.53 | 0.9561 | |
| AINDNet + TFBlind/Non-blind=Blind, Training Env.=All, Geometric Self-Ensemble=true2020.02 | 39.52 | 0.9522 | |
| VDNLearning Paradigm=Supervised2023.03 | 39.38 | 0.952 | |
| AINDNet + TFBlind/Non-blind=Blind, Training Env.=All, Geometric Self-Ensemble=false2020.02 | 39.37 | 0.9505 | |
| AINDNet(R)Blind/Non-blind=Blind, Training Env.=Real, Geometric Self-Ensemble=true2020.02 | 39.34 | 0.9524 | |
| AINDNet + RTBlind/Non-blind=Blind, Training Env.=All, Geometric Self-Ensemble=true2020.02 | 39.34 | 0.9522 | |
| AINDNet(R)Learning Paradigm=Supervised2023.03 | 39.34 | 0.952 | |
| RIDNetLearning Paradigm=Supervised2023.03 | 39.25 | 0.952 | |
| RIDNetBlind/Non-blind=Blind, Training Env.=Real2020.02 | 39.23 | 0.9526 | |
| AINDNet + RTBlind/Non-blind=Blind, Training Env.=All, Geometric Self-Ensemble=false2020.02 | 39.21 | 0.9505 | |
| AINDNet(R)Blind/Non-blind=Blind, Training Env.=Real, Geometric Self-Ensemble=false2020.02 | 39.16 | 0.9515 | |
| AT-BSN (D)Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.68 | 0.942 | |
| C-BSNLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.6 | 0.941 | |
| Spatially-Adaptive (UNet)Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.58 | 0.936 | |
| SDAP (E)Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.56 | 0.94 | |
| C-BSNLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=false2023.03 | 38.45 | 0.939 | |
| LG-BPNLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.43 | 0.942 | |
| AT-BSN (D)Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=false2023.03 | 38.29 | 0.939 | |
| AT-BSNLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.19 | 0.939 | |
| Spatially-Adaptive (UNet)Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=false2023.03 | 38.18 | 0.938 | |
| AP-BSN + R3Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 38.09 | 0.937 | |
| CBDNet(All)Blind/Non-blind=Blind, Denoising on=sRGB2018.07 | 38.06 | 0.9421 | |
| CBDNetBlind/Non-blind=Blind, Training Env.=All2020.02 | 38.06 | 0.9421 | |
| CBDNetLearning Paradigm=Supervised2023.03 | 38.05 | 0.942 | |
| TWSCBlind/Non-blind=Blind, Denoising on=sRGB2018.07 | 37.94 | 0.9403 | |
| TWSCBlind/Non-blind=Blind2020.02 | 37.94 | 0.9403 | |
| D-BSN + MWCNNLearning Paradigm=Unpaired2023.03 | 37.93 | 0.937 | |
| DnCNNLearning Paradigm=Supervised2023.03 | 37.9 | 0.943 | |
| SDAP (E)Learning Paradigm=Self-Supervised, trained_directly_on_benchmark=false2023.03 | 37.86 | 0.937 | |
| AT-BSNLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=false2023.03 | 37.76 | 0.934 | |
| CBDNet(Real)Blind/Non-blind=Blind, Denoising on=sRGB2018.07 | 37.72 | 0.9408 | |
| CBDNetBlind/Non-blind=Blind, Training Env.=Real2020.02 | 37.72 | 0.9408 | |
| CBDNet(Syn)Blind/Non-blind=Blind, Denoising on=sRGB2018.07 | 37.57 | 0.936 | |
| CBDNetBlind/Non-blind=Blind, Training Env.=Synthetic2020.02 | 37.57 | 0.936 | |
| MCWNNMBlind/Non-blind=Blind, Denoising on=sRGB2018.07 | 37.38 | 0.9294 | |
| C2NLearning Paradigm=Unpaired2023.03 | 37.28 | 0.924 | |
| CVF-SIDLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=true2023.03 | 36.5 | 0.924 | |
| KSVDBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 36.49 | 0.8978 | |
| KSVDBlind/Non-blind=Non-blind2020.02 | 36.49 | 0.8978 | |
| CVF-SIDLearning Paradigm=Self-Supervised, trained_directly_on_benchmark=false2023.03 | 36.31 | 0.923 | |
| NACLearning Paradigm=Self-Supervised2023.03 | 36.2 | 0.925 | |
| CIMMBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 36.04 | 0.9136 | |
| GCBDBlind/Non-blind=Blind, Denoising on=sRGB2018.07 | 35.58 | 0.9217 | |
| GCBDBlind/Non-blind=Blind, Training Env.=Synthetic2020.02 | 35.58 | 0.9217 | |
| GCBDLearning Paradigm=Unpaired2023.03 | 35.58 | 0.922 | |
| WNNMBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 34.67 | 0.8646 | |
| WNNMBlind/Non-blind=Non-blind2020.02 | 34.67 | 0.8646 | |
| WNNMLearning Paradigm=Non-Learning2023.03 | 34.67 | 0.865 | |
| FoEBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 34.62 | 0.8845 | |
| BM3DBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 34.51 | 0.8507 | |
| BM3DBlind/Non-blind=Non-blind2020.02 | 34.51 | 0.8507 | |
| BM3DLearning Paradigm=Non-Learning2023.03 | 34.51 | 0.851 | |
| FFDNetBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 34.4 | 0.8474 | |
| FFDNetBlind/Non-blind=Non-blind, Training Env.=Synthetic2020.02 | 34.4 | 0.8474 | |
| MLPBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 34.23 | 0.8331 | |
| MLPBlind/Non-blind=Non-blind, Training Env.=Synthetic2020.02 | 34.23 | 0.8331 | |
| NCSRBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 34.05 | 0.8351 | |
| TNRDBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 33.65 | 0.8306 | |
| TNRDBlind/Non-blind=Non-blind, Training Env.=Synthetic2020.02 | 33.65 | 0.8306 | |
| EPLLBlind/Non-blind=Non-blind, Denoising on=sRGB2018.07 | 33.51 | 0.8244 | |
| CDnCNN-BBlind/Non-blind=Blind, Denoising on=sRGB2018.07 | 32.43 | 0.79 | |
| CDnCNN-BBlind/Non-blind=Blind, Training Env.=Synthetic2020.02 | 32.43 | 0.79 |