List your company

Super-resolution on Urban100 (test)

34.7PSNRATD
26.036828.285930.53532.7841Nov 14, 2015Aug 9, 2017May 5, 2019Jan 28, 2021Oct 24, 2022Jul 19, 2024Apr 14, 2026

Evaluation results

MethodLinks
ATDScale=2x2024.1134.7———————
MambaIRv2Scale=2x2024.1134.49———————
RCANScale=x2, Kernel Width=0, Time=170ms2021.0433.34———————
DUKDScale=x2, Backbone=RCAN2023.0932.870.9349——————
RKDScale=x2, Backbone=RCAN2023.0932.70.933——————
KDScale=x2, Backbone=RCAN2023.0932.670.9329——————
FAKD*Scale=x2, Backbone=RCAN2023.0932.650.933——————
ScratchScale=x2, Backbone=RCAN2023.0932.630.9327——————
TL-Conv2026.0432.550.9337——————
B-Convgroup=p82026.0432.490.933——————
ATScale=x2, Backbone=RCAN2023.0932.480.9313——————
F-Convgroup=p82026.0432.360.9313——————
G-CNNgroup=p42026.0432.270.9306——————
CNN2026.0432.170.9297——————
FitNetScale=x2, Backbone=RCAN2023.0932.060.9279——————
Har-Netgroup=p82026.0431.990.9281——————
E2-CNNgroup=p82026.0431.990.9283——————
DASRScale=x2, Kernel Width=0, Time=71ms2021.0431.49———————
DDTBNetwork=EDSR, Scale=x2, Bit-width=4-bit2022.0631.39———————
NIPQNetwork=EDSR, Scale=x2, Bit-width=4-bit2022.0631.36———————
NIPQNetwork=EDSR, Scale=x2, Bit-width=3-bit2022.0631.13———————
TFLiteNetwork=EDSR, Scale=x2, Bit-width=4-bit2022.0631.11———————
PAMSNetwork=EDSR, Scale=x2, Bit-width=4-bit2022.0631.1———————
PACTNetwork=EDSR, Scale=x2, Bit-width=4-bit2022.0631.09———————
DDTBNetwork=EDSR, Scale=x2, Bit-width=3-bit2022.0631.01———————
PDO-eConvgroup=p82026.0430.910.9159——————
VDSRScale=x22015.1130.760.9140.98—————
DASRScale=x2, Kernel Width=0.6, Time=71ms2021.0430.71———————
PACTNetwork=EDSR, Scale=x2, Bit-width=3-bit2022.0630.57———————
TFLiteNetwork=EDSR, Scale=x2, Bit-width=3-bit2022.0630.48———————
DASRScale=x2, Kernel Width=1.2, Time=71ms2021.0430.36———————
DoReFaNetwork=EDSR, Scale=x2, Bit-width=4-bit2022.0630.17———————
SEDSR-GradResiduallearning rate=0.0003, averaging protocol=final 25% evaluations over 3 random seeds, alpha=32026.0230.06———————
DoReFaNetwork=EDSR, Scale=x2, Bit-width=3-bit2022.0630———————
EDSR-GradResiduallearning rate=0.0003, averaging protocol=final 25% evaluations over 3 random seeds, alpha=-32026.0229.99———————
EDSRlearning rate=0.0003, averaging protocol=final 25% evaluations over 3 random seeds2026.0229.93———————
SEDSRlearning rate=0.0003, averaging protocol=final 25% evaluations over 3 random seeds2026.0229.81———————
RCANScale=x2, Kernel Width=0.6, Time=170ms2021.0429.8———————
SelfExScale=x22015.1129.540.8967663.98—————
SRCNNScale=x22015.1129.50.894622.12—————
PAMSNetwork=EDSR, Scale=x2, Bit-width=3-bit2022.0629.5———————
MZSR + PredictorScale=x2, Kernel Width=0, Time=90ms2021.0429.49———————
A+Scale=x22015.1129.20.89382.96—————
RFLScale=x22015.1129.110.89043.62—————
RCANScale=x3, Kernel Width=0, Time=167ms2021.0429.09———————
SRMDNF + PredictorScale=x2, Kernel Width=0, Time=9ms2021.0429.05———————
MZSR + PredictorScale=x2, Kernel Width=0.6, Time=90ms2021.0429.01———————
DASRScale=x2, Kernel Width=1.8, Time=71ms2021.0428.95———————
DUKDScale=x3, Backbone=RCAN2023.0928.790.8646——————
KANONoise factor (N)=0, Sampling scale=22025.1228.67———————
KDScale=x3, Backbone=RCAN2023.0928.620.8612——————
FAKD*Scale=x3, Backbone=RCAN2023.0928.620.8612——————
RKDScale=x3, Backbone=RCAN2023.0928.60.861——————
ScratchScale=x3, Backbone=RCAN2023.0928.590.861——————
MZSR + PredictorScale=x2, Kernel Width=1.2, Time=90ms2021.0428.43———————
ATScale=x3, Backbone=RCAN2023.0928.430.8577——————
SRMDNF + PredictorScale=x2, Kernel Width=0.6, Time=9ms2021.0428.42———————
SRResNet-GradResidualScaling=sample-based2026.0228.27———————
FitNetScale=x3, Backbone=RCAN2023.0927.890.8472——————
DASRScale=x3, Kernel Width=0, Time=70ms2021.0427.65———————
SRResNetScaling=sample-based2026.0227.58———————
EDSRNoise factor (N)=0, Sampling scale=22025.1227.54———————
SRMDNF + PredictorScale=x2, Kernel Width=1.2, Time=9ms2021.0427.43———————
SwinIR#param=11.9M, Scale=x42023.1227.37———————
DASRScale=x3, Kernel Width=0.8, Time=70ms2021.0427.36———————
Full-ft#param=119M, Scale=x42023.1227.31———————
FacT#param=537K, Scale=x42023.1227.23———————
AdaptIR#param=370K, Scale=x42023.1227.22———————
Adapter#param=691K, Scale=x42023.1227.21———————
Adaptfor.#param=677K, Scale=x42023.1227.21———————
VPT#param=884K, Scale=x42023.1227.2———————
LORA#param=995K, Scale=x42023.1227.2———————
PretrainScale=x42023.1227.18———————
KANONoise factor (N)=5, Sampling scale=22025.1227.16———————
VDSRScale=x32015.1127.140.82791.08—————
SRMDNF + PredictorScale=x2, Kernel Width=1.8, Time=9ms2021.0427.12———————
RCANScale=x3, Kernel Width=0.8, Time=167ms2021.0426.89———————
BicubicScale=x22015.1126.880.84030—————
BicubicScale=x2, Kernel Width=02021.0426.88———————
DASRScale=x3, Kernel Width=1.6, Time=70ms2021.0426.86———————
RCAN#param=15.6M, Scale=x42023.1226.82———————
SAN#param=15.9M, Scale=x42023.1226.79———————
SRMDNF + PredictorScale=x3, Kernel Width=0, Time=6ms2021.0426.75———————
CAMixerSRScale=x4, #Params=765K, #MAdds=53.8G2024.0226.630.8012——————
DUKDScale=x4, Backbone=RCAN2023.0926.620.802——————
SRMDNF + PredictorScale=x3, Kernel Width=0.8, Time=6ms2021.0426.61———————
RCANScale=x4, Kernel Width=0, Time=165ms2021.0426.61———————
ELAN-lightScale=x4, #Params=601K, #MAdds=43.2G2024.0226.540.7982——————
SwinIR-NGScale=x4, #Params=1201K, #MAdds=63.0G2024.0226.540.7998——————
DATSR-recSR paradigm=RefSR, Loss=reconstruction loss2022.0726.520.798——————
FAKD*Scale=x4, Backbone=RCAN2023.0926.50.798——————
KDScale=x4, Backbone=RCAN2023.0926.490.798——————
SwinIR-lightScale=x4, #Params=930K, #MAdds=61.7G2024.0226.470.798——————
RKDScale=x4, Backbone=RCAN2023.0926.470.7981——————
NGswinScale=x4, #Params=1019K, #MAdds=36.4G2024.0226.450.7963——————
SelfExScale=x32015.1126.440.8088473.6—————
DiVANetScale=x4, #Params=939K, #MAdds=57.0G2024.0226.420.7958——————
KANONoise factor (N)=0, Sampling scale=32025.1226.42———————
MZSR + PredictorScale=x2, Kernel Width=1.8, Time=90ms2021.0426.39———————
ScratchScale=x4, Backbone=RCAN2023.0926.370.7949——————