Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
Published in
Neural Information Processing Systems(2017)External Links:
TL
TL;DR
The recently proposed Temporal Ensembling has achieved state-of-the-art results in several semi-supervised learning benchmarks, but it becomes unwieldy when learning large datasets, so Mean Teacher, a method that averages model weights instead of label predictions, is proposed.