English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Data augmentation (DA) is indispensable in modern machine learning and deepneural networks. The basic idea of DA is to construct new training data toimprove the model's generalization by adding slightly disturbed versions ofexisting data or synthesizing new data. In this work, we review a small butessential subset of DA -- Mix-based Data Augmentation (MixDA) that generatesnovel samples by mixing multiple examples. Unlike conventional DA approachesbased on a single-sample operation or requiring domain knowledge, MixDA is moregeneral in creating a broad spectrum of new data and has received increasingattention in the community. We begin with proposing a new taxonomy classifyingMixDA into, Mixup-based, Cutmix-based, and hybrid approaches according to ahierarchical view of the data mix. Various MixDA techniques are thencomprehensively reviewed in a more fine-grained way. Owing to itsgeneralization, MixDA has penetrated a variety of applications which are alsocompletely reviewed in this work. We also examine why MixDA works fromdifferent aspects of improving model performance, generalization, andcalibration while explaining the model behavior based on the properties ofMixDA. Finally, we recapitulate the critical findings and fundamentalchallenges of current MixDA studies, and outline the potential directions forfuture works. Different from previous related works that summarize the DAapproaches in a specific domain (e.g., images or natural language processing)or only review a part of MixDA studies, we are the first to provide asystematical survey of MixDA in terms of its taxonomy, methodology,applications, and explainability. This work can serve as a roadmap to MixDAtechniques and application reviews while providing promising directions forresearchers interested in this exciting area.

A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainability