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

Emotion recognition is a complex task due to the inherent subjectivity inboth the perception and production of emotions. The subjectivity of emotionsposes significant challenges in developing accurate and robust computationalmodels. This thesis examines critical facets of emotion recognition, beginningwith the collection of diverse datasets that account for psychological factorsin emotion production.  To handle the challenge of non-representative training data, this workcollects the Multimodal Stressed Emotion dataset, which introduces controlledstressors during data collection to better represent real-world influences onemotion production. To address issues with label subjectivity, this researchcomprehensively analyzes how data augmentation techniques and annotationschemes impact emotion perception and annotator labels. It further handlesnatural confounding variables and variations by employing adversarial networksto isolate key factors like stress from learned emotion representations duringmodel training. For tackling concerns about leakage of sensitive demographicvariables, this work leverages adversarial learning to strip sensitivedemographic information from multimodal encodings. Additionally, it proposesoptimized sociological evaluation metrics aligned with cost-effective,real-world needs for model testing.  This research advances robust, practical emotion recognition throughmultifaceted studies of challenges in datasets, labels, modeling, demographicand membership variable encoding in representations, and evaluation. Thegroundwork has been laid for cost-effective, generalizable emotion recognitionmodels that are less likely to encode sensitive demographic information.

Implicit Design Choices and Their Impact on Emotion Recognition Model Development and Evaluation