Image Classification on MNIST -> USPS (test)
99.5AccuracyDTA
Evaluation Results
| Method | Links | |
|---|---|---|
| DTA2019.10 | 99.5 | |
| Mean TeacherTraining Protocol=Unsupervised Domain Adaptation, Augmentation Type=TF2017.06 | 98.26 | |
| Mean TeacherTraining Protocol=Unsupervised Domain Adaptation, Augmentation Type=TFA, Confidence Thresholding=true2017.06 | 98.23 | |
| Mean TeacherTraining Protocol=Unsupervised Domain Adaptation, Augmentation Type=TF, Confidence Thresholding=true2017.06 | 98.13 | |
| SE*Protocol=MT+CT+TF2019.10 | 98.1 | |
| SupTgtTraining Protocol=Train on Target, Augmentation Type=TFA2017.06 | 97.83 | |
| Target only (baseline)training=Target domain only2019.10 | 97.8 | |
| SupTgtTraining Protocol=Train on Target, Augmentation Type=TF2017.06 | 97.65 | |
| SupTgtTraining Protocol=Train on Target, Augmentation Type=minimal2017.06 | 97.29 | |
| MCD (Ours)n (number of feature generator updates)=42017.12 | 96.5 | |
| Proposed MethodTask=Rotation2019.09 | 96.5 | |
| Source only (baseline)training=Source domain only2019.10 | 96.3 | |
| SupSrcTraining Protocol=Train on Source, Augmentation Type=TFA2017.06 | 96.25 | |
| PixelDA2017.12 | 95.9 | |
| MCD (Ours)n (number of feature generator updates)=32017.12 | 95.6 | |
| CyCADA2019.09 | 95.6 | |
| SupSrcTraining Protocol=Train on Source, Augmentation Type=TF2017.06 | 95.39 | |
| SBADA-GANTraining Protocol=Unsupervised Domain Adaptation2017.06 | 95.04 | |
| Source only (Our method)mode=source only2019.09 | 94.7 | |
| MCD (Ours)n (number of feature generator updates)=42017.12 | 94.2 | |
| MCD (Ours)n (number of feature generator updates)=32017.12 | 93.8 | |
| MCD (Ours)n (number of feature generator updates)=22017.12 | 93.1 | |
| G2ATraining Protocol=Unsupervised Domain Adaptation2017.06 | 92.5 | |
| MCD (Ours)n (number of feature generator updates)=22017.12 | 92.1 | |
| DRCN2017.12 | 91.8 | |
| DCRNTraining Protocol=Unsupervised Domain Adaptation2017.06 | 91.8 | |
| DSNlabeled target samples for validation=true2017.12 | 91.3 | |
| CoGAN2017.02 | 91.2 | |
| COGAN2017.12 | 91.2 | |
| RevGradTraining Protocol=Unsupervised Domain Adaptation2017.06 | 91.11 | |
| SWB1DGk=1002025.10 | 90.1 | |
| SOTDDk=1002025.10 | 89.6 | |
| ADDA2017.02 | 89.4 | |
| ADDA2017.12 | 89.4 | |
| ADDATraining Protocol=Unsupervised Domain Adaptation2017.06 | 89.4 | |
| ADDA2019.09 | 89.4 | |
| COGAN2017.12 | 89.1 | |
| Mean TeacherTraining Protocol=Unsupervised Domain Adaptation, Augmentation Type=minimal, Confidence Thresholding=true2017.06 | 88.14 | |
| Trained on Qk=1002025.10 | 87.5 | |
| DANNlabeled target samples for validation=true2017.12 | 85.1 | |
| SupSrcTraining Protocol=Train on Source, Augmentation Type=minimal2017.06 | 82.03 | |
| MMD2017.12 | 81.1 | |
| JGSAkernel=primal2017.05 | 80.44 | |
| SOTDDk=102025.10 | 80.1 | |
| Source Onlyadaptation=none2017.12 | 79.4 | |
| Domain confusion2017.02 | 79.1 | |
| SWB1DGk=102025.10 | 77.8 | |
| Gradient reversal2017.02 | 77.1 | |
| DANNlabeled target samples for validation=true2017.12 | 77.1 | |
| Source Onlyadaptation=none2017.12 | 76.7 | |
| SOTDDk=52025.10 | 75.7 | |
| Source only2017.02 | 75.2 | |
| SWB1DGk=52025.10 | 73.5 | |
| OTDDk=52025.10 | 73.3 | |
| OTDDk=102025.10 | 72.7 | |
| SA2017.05 | 67.78 | |
| JDA2017.05 | 67.28 | |
| Rawadaptation=none2017.05 | 65.94 | |
| SCABase Classifier=1NN, Supervision=Semi-supervised2015.10 | 65.11 | |
| SCA2017.05 | 65.11 | |
| SDA2017.05 | 65 | |
| uSCABase Classifier=1NN, Supervision=Unsupervised2015.10 | 64.67 | |
| SABase Classifier=1NN2015.10 | 63.95 | |
| TJM2017.05 | 63.28 | |
| RawBase Classifier=1NN2015.10 | 63.06 | |
| KPCABase Classifier=1NN2015.10 | 62.61 | |
| TJMBase Classifier=1NN2015.10 | 62 | |
| GFKBase Classifier=1NN2015.10 | 61.22 | |
| GFK2017.05 | 61.22 | |
| Trained on Qk=102025.10 | 60.3 | |
| SSTCABase Classifier=1NN2015.10 | 60.13 | |
| TSCBase Classifier=1NN2015.10 | 59.56 | |
| TCABase Classifier=1NN2015.10 | 59.44 | |
| TCA2017.05 | 56.33 | |
| Trained on Qk=52025.10 | 51.4 | |
| SOTDDk=12025.10 | 50.1 | |
| SWB1DGk=12025.10 | 45.4 | |
| OTDDk=12025.10 | 39.5 | |
| Trained on Qk=12025.10 | 32.4 |