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/

Zendy Plus

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

Zendy Tools

Zendy Open

Data mixing augmentation have proved to be effective in improving thegeneralization ability of deep neural networks. While early methods mix samplesby hand-crafted policies (e.g., linear interpolation), recent methods utilizesaliency information to match the mixed samples and labels via complex offlineoptimization. However, there arises a trade-off between precise mixing policiesand optimization complexity. To address this challenge, we propose a novelautomatic mixup (AutoMix) framework, where the mixup policy is parameterizedand serves the ultimate classification goal directly. Specifically, AutoMixreformulates the mixup classification into two sub-tasks (i.e., mixed samplegeneration and mixup classification) with corresponding sub-networks and solvesthem in a bi-level optimization framework. For the generation, a learnablelightweight mixup generator, Mix Block, is designed to generate mixed samplesby modeling patch-wise relationships under the direct supervision of thecorresponding mixed labels. To prevent the degradation and instability ofbi-level optimization, we further introduce a momentum pipeline to trainAutoMix in an end-to-end manner. Extensive experiments on nine image benchmarksprove the superiority of AutoMix compared with state-of-the-art in variousclassification scenarios and downstream tasks.

AutoMix: Unveiling the Power of Mixup for Stronger Classifiers