Image Classification on ImageNetV2 Top Images Ti (test)
79.4Top-1 AccRELICv2
Evaluation Results
| Method | Links | |
|---|---|---|
| RELICv2Protocol=Linear evaluation, Backbone=ResNet-50, Pre-training=ImageNet, Evaluation=Zero-shot2022.01 | 79.4 | |
| SupervisedProtocol=Linear evaluation, Backbone=ResNet-50, Pre-training=ImageNet, Evaluation=Zero-shot2022.01 | 78.4 | |
| RELICProtocol=Linear evaluation, Backbone=ResNet-50, Pre-training=ImageNet, Evaluation=Zero-shot2022.01 | 77.7 | |
| BYOLProtocol=Linear evaluation, Backbone=ResNet-50, Pre-training=ImageNet, Evaluation=Zero-shot2022.01 | 77 | |
| RELICv2Epochs=3000, Pre-training Dataset=JFT-300M, Evaluation Protocol=linear evaluation2022.01 | 73.4 | |
| RELICv2Epochs=5000, Pre-training Dataset=JFT-300M, Evaluation Protocol=linear evaluation2022.01 | 73.3 | |
| RELICv2Epochs=1000, Pre-training Dataset=JFT-300M, Evaluation Protocol=linear evaluation2022.01 | 73 | |
| SimCLRProtocol=Linear evaluation, Backbone=ResNet-50, Pre-training=ImageNet, Evaluation=Zero-shot2022.01 | 68 |