FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence

Open paper Other version
Benchmark Model Rank Results
image-classification-on-stl-10FixMatch (CTA)#12Percentage correct: 94.83
image-classification-on-stl-10ReMixMatch#13Percentage correct: 94.77
image-classification-on-stl-10UDA#20Percentage correct: 92.34
image-classification-on-stl-10FixMatch (RA)#22Percentage correct: 92.02
image-classification-on-stl-10MixMatch#29Percentage correct: 89.59
image-classification-on-stl-10Mean Teacher#44Percentage correct: 78.57
image-classification-on-stl-10Π-Model#56Percentage correct: 73.77
image-classification-on-stl-10Pseudo-Labeling#58Percentage correct: 72.01
semi-supervised-image-classification-on-2FixMatch#49Top 5 Accuracy: 89.13%
semi-supervised-image-classification-on-cifarFixMatch (CTA)#9Percentage error: 4.31
semi-supervised-image-classification-on-cifar-2FixMatch (RA, WRN-28-8)#9Percentage error: 22.6
semi-supervised-image-classification-on-cifar-2FixMatch (CTA, WRN-28-8)#12Percentage error: 23.18±0.11
semi-supervised-image-classification-on-cifar-36FixMatch#3Accuracy: 56.8
semi-supervised-image-classification-on-cifar-37FixMatch#3Accuracy: 70.2
semi-supervised-image-classification-on-cifar-38FixMatch#3Accuracy: 83.7
semi-supervised-image-classification-on-cifar-6FixMatch (CTA)#8Percentage error: 5.07±0.33
semi-supervised-image-classification-on-cifar-7FixMatch (CTA)#13Percentage error: 11.39±3.35
semi-supervised-image-classification-on-cifar-8FixMatch (CTA)#14Percentage error: 49.95±3.01
semi-supervised-image-classification-on-cifar-9FixMatch (CTA, WRN-28-8)#10Percentage error: 28.64±0.24
semi-supervised-image-classification-on-stl-1FixMatch (CTA)#3Accuracy: 94.83±0.63
semi-supervised-image-classification-on-svhnFixMatch (CTA)#3Accuracy: 97.64±0.19
semi-supervised-image-classification-on-svhn-2FixMatch (CTA)#4Percentage error: 7.65±7.65

Rank counts only results with a code link.