MixMatch: A Holistic Approach to Semi-Supervised Learning

Benchmark Model Rank Results
semi-supervised-image-classification-on-cifar-10-1000-labelsMixMatch#1Accuracy: 92.25
semi-supervised-image-classification-on-stl-10-1000-labelsMixMatch#7Accuracy: 89.82
semi-supervised-image-classification-on-svhn-1000-labelsMixMatch#7Accuracy: 96.73
semi-supervised-image-classification-on-svhn-250-labelsMixMatch#8Accuracy: 96.22
semi-supervised-image-classification-on-cifar-10-250-labelsMixMatch#14Percentage error: 11.08
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
semi-supervised-image-classification-on-cifar-10-4000-labelsMixMatch#21Percentage error: 6.24
image-classification-on-svhnMixMatch#31Percentage error: 2.59
image-classification-on-cifar-10MixMatch#168Percentage correct: 95.05
image-classification-on-cifar-100MixMatch#192Percentage correct: 74.1