google-research/uda
Unsupervised Data Augmentation (UDA) is a semi-supervised learning technique from Google Research that achieves state-of-the-art results on text and image classification.
Not currently ranked — collecting fresh signals.
star history
UDA is a semi-supervised learning method that reduces the need for labeled examples by better utilizing unlabeled ones. The repository provides TensorFlow code for text classification using BERT and image classification on CIFAR-10 and SVHN, achieving significant error rate reductions over prior methods. It generates augmented examples through back-translation for text and various augmentation techniques for images.
Frequently asked
- What is google-research/uda?
- Unsupervised Data Augmentation (UDA) is a semi-supervised learning technique from Google Research that achieves state-of-the-art results on text and image classification.
- Is uda open source?
- Yes — google-research/uda is open source, released under the Apache-2.0 license.
- What language is uda written in?
- google-research/uda is primarily written in Python.
- How popular is uda?
- google-research/uda has 2.2k stars on GitHub.
- Where can I find uda?
- google-research/uda is on GitHub at https://github.com/google-research/uda.