Probabilistic Contrastive Learning for Domain Adaptation

Benchmark Model Rank Results
semi-supervised-image-classification-on-cifar-100-400-labelsPCL (Flexmatch)#5Percentage error: 35.75±0.53
semi-supervised-image-classification-on-cifar-100-400-labelsPCL (Fixmatch)#12Percentage error: 42.38±2.52
semi-supervised-image-classification-on-cifar-10-40-labelsPCL#10Percentage error: 5.88±1.19
domain-adaptation-on-office-homePCL#11Accuracy: 74.5