Large Scale Adversarial Representation Learning

Open paper
Benchmark Model Rank Results
contrastive-learning-on-imagenet-1kResNet50 (4×)#10ImageNet Top-1 Accuracy: 61.3
self-supervised-image-classification-on-imagenetBigBiGAN (RevNet-50 ×4, BN+CReLU)#113Top 1 Accuracy: 61.3%Top 5 Accuracy: 81.9%Number of Params: 86M
self-supervised-image-classification-on-imagenetBigBiGAN (RevNet-50 ×4)#115Top 1 Accuracy: 60.8%Top 5 Accuracy: 81.4%Number of Params: 86M
self-supervised-image-classification-on-imagenetBigBiGAN (ResNet-50, BN+CReLU)#121Top 1 Accuracy: 56.6%Top 5 Accuracy: 78.6%Number of Params: 24M
self-supervised-image-classification-on-imagenetBigBiGAN (ResNet-50)#124Top 1 Accuracy: 55.4%Top 5 Accuracy: 77.4%Number of Params: 25M
semi-supervised-image-classification-on-1BigBiGAN (RevNet-50 ×4, BN+CReLU)#51Top 5 Accuracy: 55.2%
semi-supervised-image-classification-on-2BigBiGAN (RevNet-50 ×4, BN+CReLU)#65Top 5 Accuracy: 78.8%

Rank counts only results with a code link.