In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning

Benchmark Model Rank Results
semi-supervised-image-classification-on-cifar-10-1000-labelsUPS (CNN-13)#2Accuracy: 91.82
semi-supervised-medical-image-classification-on-chest-x-ray14-2-labeledUPS#3AUC: 65.51
semi-supervised-image-classification-on-cifar-10-4000-labelsUPS (Shake-Shake)#12Percentage error: 4.86
semi-supervised-image-classification-on-cifar-10-4000-labelsUPS (CNN-13)#24Percentage error: 6.39±0.02
semi-supervised-image-classification-on-cifar-100-10000-labelsUPS (CNN-13)#16Percentage error: 32