List your company

4x Super-Resolution on DIV2K (val)

30.15PSNRDSRNet
26.062827.123928.18529.2461Jan 13, 2022May 25, 2022Oct 4, 2022Feb 13, 2023Jun 25, 2023Nov 4, 2023Mar 16, 2024

Evaluation results

MethodLinks
DSRNetScale=x42023.1030.15—————
EDSR+Scale=x42023.1029.38—————
MDSR+Scale=x42023.1029.36—————
MDSRScale=x42023.1029.26—————
EDSRScale=x42023.1029.25—————
FxSR-PDParams=18.3M, Condition Parameter (t)=0.02022.0129.240.838353.30.2390.11694.11
SRResNetScale=x4, Loss Function=L2 loss2023.1028.92—————
SRResNetScale=x4, Loss Function=L1 loss2023.1028.92—————
VDSRScale=x42023.1028.17—————
RRDB-LINF-LPCategory=Flow-based Generative SR Method, Backbone=RRDB, Upsampling scale=4x2024.03280.7847.30.105——
SRCNNScale=x42023.1027.78—————
A+Scale=x42023.1027.7—————
SROOECategory=Generative SR Method, Upsampling scale=4x2024.0327.690.7950.80.096——
EDSR-baseline-LINF-LPCategory=Flow-based Generative SR Method, Backbone=EDSR-baseline, Upsampling scale=4x2024.0327.640.7846.960.119——
FxSR-PDParams=18.3M, Condition Parameter (t)=0.82022.0127.510.78950.540.10280.05132.81
SRFlow-LPCategory=Flow-based Generative SR Method, Upsampling scale=4x2024.0327.510.7851.510.109——
SRDiffCategory=Generative SR Method, Upsampling scale=4x2024.0327.410.7955.210.136——
RRDB-LINFCategory=Flow-based Generative SR Method, Sampling temperature (T)=0.8, Backbone=RRDB, Upsampling scale=4x2024.0327.330.7743.640.112——
SRFlowCategory=Flow-based Generative SR Method, Sampling temperature (T)=0.9, Upsampling scale=4x2024.0327.090.7649.960.121——
EDSR-baseline-LINFCategory=Flow-based Generative SR Method, Sampling temperature (T)=0.8, Backbone=EDSR-baseline, Upsampling scale=4x2024.0327.020.7643.190.13——
BicubicScale=x42023.1026.66—————
HCFlow++Category=Flow-based Generative SR Method, Sampling temperature (T)=0.9, Upsampling scale=4x2024.0326.610.7450.070.11——
RankSRGANCategory=Generative SR Method, Upsampling scale=4x2024.0326.550.7542.330.128——
ESRGANCategory=Generative SR Method, Upsampling scale=4x2024.0326.220.7539.030.124——