Diversity Helps: Unsupervised Few-shot Learning via Distribution Shift-based Data Augmentation
About
Few-shot learning aims to learn a new concept when only a few training examples are available, which has been extensively explored in recent years. However, most of the current works heavily rely on a large-scale labeled auxiliary set to train their models in an episodic-training paradigm. Such a kind of supervised setting basically limits the widespread use of few-shot learning algorithms. Instead, in this paper, we develop a novel framework called Unsupervised Few-shot Learning via Distribution Shift-based Data Augmentation (ULDA), which pays attention to the distribution diversity inside each constructed pretext few-shot task when using data augmentation. Importantly, we highlight the value and importance of the distribution diversity in the augmentation-based pretext few-shot tasks, which can effectively alleviate the overfitting problem and make the few-shot model learn more robust feature representations. In ULDA, we systemically investigate the effects of different augmentation techniques and propose to strengthen the distribution diversity (or difference) between the query set and support set in each few-shot task, by augmenting these two sets diversely (i.e., distribution shifting). In this way, even incorporated with simple augmentation techniques (e.g., random crop, color jittering, or rotation), our ULDA can produce a significant improvement. In the experiments, few-shot models learned by ULDA can achieve superior generalization performance and obtain state-of-the-art results in a variety of established few-shot learning tasks on Omniglot and miniImageNet. The source code is available in https://github.com/WonderSeven/ULDA.
Key insights
What problem does ULDA address?
Unsupervised few-shot learning trains models without labeled auxiliary data. Prior augmentation methods make query and support sets too similar, causing overfitting. ULDA augments them differently to create distribution shift, producing harder tasks and more robust features.
How does ULDA differ from prior work?
Unlike AAL and UMTRA, which use identical augmentation for both sets, ULDA applies separate operators. On miniImageNet 5-way 1-shot, ULDA-ProtoNets reaches 40.63% versus CACTUs-ProtoNets at 39.18%, evaluated on Omniglot, miniImageNet, and tieredImageNet.
What are the practical uses?
ULDA helps train few-shot classifiers when labels are scarce or expensive, such as domain-professional areas.
What limitations do the authors note?
With deeper backbones (ResNet34, ResNet50), performance drops due to a new overfitting risk; the authors flag increasing task diversity as future work. Evaluation is limited to image classification benchmarks.