SelfMatch: Combining Contrastive Self-Supervision and Consistency for Semi-Supervised Learning

Benchmark Model Rank Results
semi-supervised-image-classification-on-cifar-10-250-labelsSelfMatch–Percentage error: 4.87±0.26
semi-supervised-image-classification-on-cifar-10-40-labelsSelfMatch–Percentage error: 6.81±1.08
semi-supervised-image-classification-on-cifar-10-4000-labelsSelfMatch–Percentage error: 4.06±0.08