Review -- UDA: Unsupervised Data Augmentation for Consistency Training

This validates the idea of stronger data augmentations found in supervised learning can always lead to more gains when applied to the semi-supervised learning settings. First, UDA consistently outperforms the two baselines given different sizes of labeled data. Moreover, the performance difference between UDA and VAT shows the superiority of data augmentation based noise. Given the same architecture, UDA outperforms all published results by significant margins and nearly matches the fully supervised performance, which uses 10 more labeled examples. First, even with very few labeled examples, UDA can offer decent or even competitive performances compared to the SOTA model trained with full supervised data.