List your company
ResearchFormerly WizWand

Digit Classification on MNIST -> USPS (test)

98.5AccuracyDeepJDOT
30.099247.857165.61583.3729Mar 27, 2018Nov 6, 2018Jun 18, 2019Jan 28, 2020Sep 8, 2020Apr 20, 2021Dec 1, 2021
Updated 8mo ago

Evaluation results

MethodLinks
DeepJDOTadaptation=source to target, eval_domain=source2018.0398.5
Target Supervisedtraining=full supervised2019.0996.3
3CATN2019.0996.1
UNIT2019.0995.9
UNITadaptation=source to target2018.0395.9
IntraDAArchitecture=LeNet variant2020.0495.8
target onlyadaptation=none, train_domain=target2018.0395.8
DeepJDOTBackbone=LeNet2021.0795.7
DANNadaptation=source to target2018.0395.7
DeepJDOTadaptation=source to target2018.0395.7
CDANvariant=CDAN+E2019.0995.6
CyCADAArchitecture=LeNet variant2020.0495.6
GenToAdaptadaptation=source to target2018.0395.3
PFAN2018.1195
Fully SupervisedArchitecture=Simple CNN (2 conv, 2 fc, BN)2021.1295
Source onlyadaptation=none, train_domain=source2018.0394.8
SRDA-Gnoise_method=Gaussian, image_level_optimization=false2019.0594.76
MCD2019.0594.2
StochJDOTadaptation=source to target2018.0393.6
SRDA-G*noise_method=Gaussian, image_level_optimization=true2019.0593.25
MSTN2018.1192.9
ADDAadaptation=source to target2018.0392.4
I2I Adaptadaptation=source to target2018.0392.1
DRCN2019.0991.8
DRCN2019.0591.8
DRCNadaptation=source to target2018.0391.8
DSN2019.0591.3
DSNadaptation=source to target2018.0391.3
CoGAN2019.0991.2
COGAN2019.0591.2
CoGANadaptation=source to target2018.0391.2
TOHANNumber of Target Data per Class=72021.0690.4
TOHANNumber of Target Data per Class=62021.0690
ADDA2018.1189.4
ADDA2019.0989.4
ADDAArchitecture=LeNet variant2020.0489.4
TOHANNumber of Target Data per Class=52021.0689.4
ADDA2019.0589.4
DeepCORALadaptation=source to target2018.0389.33
TOHANNumber of Target Data per Class=42021.0689.3
RevGrad2019.0989.1
TOHANNumber of Target Data per Class=32021.0688.5
MMDadaptation=source to target2018.0388.5
SRDA-V*noise_method=VAT, image_level_optimization=true2019.0588.49
TOHANNumber of Target Data per Class=22021.0688.3
Self-ensembleadaptation=source to target2018.0388.14
TOHANNumber of Target Data per Class=12021.0687.7
DUAArchitecture=Simple CNN (2 conv, 2 fc, BN)2021.1286
SRDA-Fnoise_method=FGSM, image_level_optimization=false2019.0585
SRDA-Vnoise_method=VAT, image_level_optimization=false2019.0584.64
Source onlytraining=source only2019.0982.2
Source onlyArchitecture=LeNet variant2020.0482.2
dkdHTLNumber of Target Data per Class=72021.0679
dkdHTLNumber of Target Data per Class=62021.0678.8
dkdHTLNumber of Target Data per Class=52021.0678.6
Source onlyArchitecture=Simple CNN (2 conv, 2 fc, BN)2021.1278
dkdHTLNumber of Target Data per Class=42021.0677.8
RevGrad2018.1177.1
DANN2019.0577.1
Source Onlyadaptation_protocol=No adaptation2019.0576.7
Source Only2018.1175.2
dkdHTLNumber of Target Data per Class=32021.0674.4
dkdHTLNumber of Target Data per Class=22021.0670.5
dkdHTLNumber of Target Data per Class=12021.0665.2
SRDA-F*noise_method=FGSM, image_level_optimization=true2019.0532.73