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

Wearables are fundamental to improving our understanding of human activities,especially for an increasing number of healthcare applications fromrehabilitation to fine-grained gait analysis. Although our collective know-howto solve Human Activity Recognition (HAR) problems with wearables hasprogressed immensely with end-to-end deep learning paradigms, severalfundamental opportunities remain overlooked. We rigorously explore these newopportunities to learn enriched and highly discriminating activityrepresentations. We propose: i) learning to exploit the latent relationshipsbetween multi-channel sensor modalities and specific activities; ii)investigating the effectiveness of data-agnostic augmentation for multi-modalsensor data streams to regularize deep HAR models; and iii) incorporating aclassification loss criterion to encourage minimal intra-class representationdifferences whilst maximising inter-class differences to achieve morediscriminative features. Our contributions achieves new state-of-the-artperformance on four diverse activity recognition problem benchmarks with largemargins -- with up to 6% relative margin improvement. We extensively validatethe contributions from our design concepts through extensive experiments,including activity misalignment measures, ablation studies and insights sharedthrough both quantitative and qualitative studies.

Attend And Discriminate: Beyond the State-of-the-Art for Human Activity Recognition using Wearable Sensors