MixMatch: A Holistic Approach to Semi-Supervised Learning

Benchmark Model Rank Results
image-classification-on-cifar-10MixMatch#137Percentage correct: 95.05
image-classification-on-cifar-100MixMatch#148Percentage correct: 74.1
image-classification-on-stl-10MixMatch#15Percentage correct: 94.41
image-classification-on-stl-10IIC#34Percentage correct: 88.80
image-classification-on-stl-10CutOut#36Percentage correct: 87.36
image-classification-on-svhnMixMatch#31Percentage error: 2.59
semi-supervised-image-classification-on-cifarMixMatch#21Percentage error: 6.24
semi-supervised-image-classification-on-cifar-11MixMatch#1Accuracy: 92.25
semi-supervised-image-classification-on-cifar-6MixMatch#14Percentage error: 11.08
semi-supervised-image-classification-on-stl-1MixMatch#7Accuracy: 89.82
semi-supervised-image-classification-on-svhnMixMatch#7Accuracy: 96.73
semi-supervised-image-classification-on-svhn-1MixMatch#8Accuracy: 96.22