Meta Pseudo Labels

Open paper
Benchmark Model Rank Results
semi-supervised-image-classification-on-svhn-1000-labelsMeta Pseudo Labels (WRN-28-2)#1Accuracy: 98.01 ± 0.07
semi-supervised-image-classification-on-cifar-10-4000-labelsMeta Pseudo Labels (WRN-28-2)#2Percentage error: 3.89± 0.07
image-classification-on-imagenet-realMeta Pseudo Labels (EfficientNet-B6-Wide)#4Accuracy: 91.12%
image-classification-on-imagenet-realMeta Pseudo Labels (EfficientNet-L2)#9Accuracy: 91.02%
image-classification-on-imagenetMeta Pseudo Labels (EfficientNet-L2)#6Top 1 Accuracy: 90.2%Number of params: 480MHardware Burden: 95040G…
image-classification-on-imagenetMeta Pseudo Labels (EfficientNet-B6-Wide)#12Top 1 Accuracy: 90%Number of params: 390M
image-classification-on-imagenetMeta Pseudo Labels (ResNet-50)#480Top 1 Accuracy: 83.2%
semi-supervised-image-classification-on-imagenet-10-labeled-dataMeta Pseudo Labels (ResNet-50)#29Top 1 Accuracy: 73.89%Top 5 Accuracy: 91.38%

Rank counts only results with a code link.