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

Stress is a complex issue with wide-ranging physical and psychologicalimpacts on human daily performance. Specifically, acute stress detection isbecoming a valuable application in contextual human understanding. Two commonapproaches to training a stress detection model are subject-dependent andsubject-independent training methods. Although subject-dependent trainingmethods have proven to be the most accurate approach to build stress detectionmodels, subject-independent models are a more practical and cost-efficientmethod, as they allow for the deployment of stress level detection andmanagement systems in consumer-grade wearable devices without requiringtraining data for the end-user. To improve the performance ofsubject-independent stress detection models, in this paper, we introduce astress-related bio-signal processing pipeline with a simple neural networkarchitecture using statistical features extracted from multimodal contextualsensing sources including Electrodermal Activity (EDA), Blood Volume Pulse(BVP), and Skin Temperature (ST) captured from a consumer-grade wearabledevice. Using our proposed model architecture, we compare the accuracy betweenstress detection models that use measures from each individual signal source,and one model employing the fusion of multiple sensor sources. Extensiveexperiments on the publicly available WESAD dataset demonstrate that ourproposed model outperforms conventional methods as well as providing 1.63%higher mean accuracy score compared to the state-of-the-art model whilemaintaining a low standard deviation. Our experiments also show that combiningfeatures from multiple sources produce more accurate predictions than usingonly one sensor source individually.

An Improved Subject-Independent Stress Detection Model Applied to Consumer-grade Wearable Devices