Image Classification on ImageNet 100% labels
96.5Top-5 AccuracyCPC v2
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CPC v2Method Category=Representation learning, Evaluation Protocol=Fine-tuned2019.05 | 96.5 | — | |
| CPC v2Method Category=Representation learning, Evaluation Protocol=Fixed2019.05 | 96.2 | — | |
| SimCLRArchitecture=ResNet-50 (4x), Evaluation Protocol=Fine-tuning, Data Augmentation=Broader composition2020.02 | 95.4 | 80.4 | |
| Supervised BaselineMethod Category=Supervised2019.05 | 95.2 | — | |
| Rot. + VAT + Ent. Min.Method Category=Label-propagation2019.05 | 95 | — | |
| SimCLRArchitecture=ResNet-50 (2x), Evaluation Protocol=Fine-tuning, Data Augmentation=Broader composition2020.02 | 94.8 | 79.1 | |
| UDA (RandAugment)SSL=true, Backbone=ResNet-50, Image size=3312019.04 | 94.49 | 79.05 | |
| ResNet-50 w. RandAugmentSSL=false, Backbone=ResNet-50, Image size=3312019.04 | 94.37 | 78.43 | |
| ResNet-50 (4x)Architecture=ResNet-50 (4x), Evaluation Protocol=Trained from scratch2020.02 | 94.2 | 78.4 | |
| ResNet-50SSL=false, Backbone=ResNet-50, Image size=3312019.04 | 93.73 | 77.28 | |
| SimCLRArchitecture=ResNet-50, Evaluation Protocol=Fine-tuning, Data Augmentation=Broader composition2020.02 | 93.1 | 76 | |
| AMDIMMethod Category=Representation learning, Evaluation Protocol=Fixed2019.05 | 92.2 | — | |
| BigBiGANMethod Category=Representation learning, Evaluation Protocol=Fixed2019.05 | 87 | — |