List your company
ResearchFormerly WizWand

Domain Adaptation on SVHN to MNIST (test)

98.8AccuracyrRevGrad+CAT
55.95267.07678.289.324Nov 7, 2016Jul 27, 2017Apr 16, 2018Jan 3, 2019Sep 23, 2019Jun 11, 2020Mar 1, 2021
Updated 8mo ago

Evaluation results

MethodLinks
rRevGrad+CAT2019.0398.8
CAT2019.0398.1
RevGrad+CAT2019.0398
DWTBackbone=LeNet2019.0997.7
ADABackbone=LeNet2019.0997.6
AssocDA2019.0397.6
MCD+CAT2019.0397.1
MCD2019.0396.2
Inference-time Label-Preserving Target ProjectionsLabeled target samples (T)=102021.0395.33
MEGABackbone=LeNet2019.0995.2
VADA+CAT2019.0395.2
CCSALabeled target samples (T)=102021.0394.57
VADA2019.0394.5
Inference-time Label-Preserving Target ProjectionsLabeled target samples (T)=72021.0392.02
M-ADALabeled target samples (T)=102021.0391.81
JAENumber of target samples per digit=20 samples2017.0591.8
MSTN2019.0391.7
M-ADALabeled target samples (T)=72021.0389.9
AutoDIALBackbone=LeNet2019.0989.12
FADALabeled target samples (T)=72021.0387.2
Asymmetric Tri-trainingWeight constraint=lambda = 02017.0286.2
ATTBackbone=LeNet2019.0986.2
Asymmetric Tri-training2017.0285
DSN2017.0282.7
DRCN2017.0282
DRCNBackbone=LeNet2019.0982
DRCN2019.0382
LEL2019.0381
JAENumber of target samples per digit=5 samples2017.0580.5
Asymmetric Tri-trainingBatch Normalization=false2017.0279.8
DTNsource_split=test split2016.1179.72
KNN-Ad2017.0278.8
SBADA-GANBackbone=LeNet2019.0976.1
ADDABackbone=LeNet2019.0976
ADDA2019.0376
ADDA2017.0576
GAMBackbone=LeNet2019.0974.6
DANNBackbone=LeNet2019.0973.9
RevGrad2019.0373.9
Gradient reversal2017.0573.9
DANN2016.1173.85
Shape-Biased FrameworkBackbone=LeNet2019.0971.7
MMD2017.0271.1
DANN2017.0271.1
Source Only with BNBatch Normalization=true2017.0270.1
Source Only w/o BNBatch Normalization=false2017.0268.1
DDC2019.0368.1
Domain confusion2017.0568.1
Inference-time Label-Preserving Target ProjectionsLabeled target samples (T)=02021.0364.12
M-ADALabeled target samples (T)=02021.0360.14
Source Only2019.0360.1
SA2016.1159.32
JiGenBackbone=LeNet2019.0957.6