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

Mixup style data augmentation algorithms have been widely adopted in varioustasks as implicit network regularization on representation learning to improvemodel generalization, which can be achieved by a linear interpolation oflabeled samples in input or feature space as well as target space. Inspired bygood robustness of alternative dropout strategies against over-fitting onlimited patterns of training samples, this paper introduces a novel concept ofShuffleMix -- Shuffle of Mixed hidden features, which can be interpreted as akind of dropout operation in feature space. Specifically, our ShuffleMix methodfavors a simple linear shuffle of randomly selected feature channels forfeature mixup in-between training samples to leverage semantic interpolatedsupervision signals, which can be extended to a generalized shuffle operationvia additionally combining linear interpolations of intra-channel features.Compared to its direct competitor of feature augmentation -- the ManifoldMixup, the proposed ShuffleMix can gain superior generalization, owing toimposing more flexible and smooth constraints on generating samples andachieving regularization effects of channel-wise feature dropout. Experimentalresults on several public benchmarking datasets of single-label and multi-labelvisual classification tasks can confirm the effectiveness of our method onconsistently improving representations over the state-of-the-art mixupaugmentation.

ShuffleMix: Improving Representations via Channel-Wise Shuffle of Interpolated Hidden States