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

Multimodal emotion recognition in conversations (mERC) is an active researchtopic in natural language processing (NLP), which aims to predict human'semotional states in communications of multiple modalities, e,g., naturallanguage and facial gestures. Innumerable implicit prejudices andpreconceptions fill human language and conversations, leading to the questionof whether the current data-driven mERC approaches produce a biased error. Forexample, such approaches may offer higher emotional scores on the utterances byfemales than males. In addition, the existing debias models mainly focus ongender or race, where multibias mitigation is still an unexplored task in mERC.In this work, we take the first step to solve these issues by proposing aseries of approaches to mitigate five typical kinds of bias in textualutterances (i.e., gender, age, race, religion and LGBTQ+) and visualrepresentations (i.e, gender and age), followed by a Multibias-Mitigated andsentiment Knowledge Enriched bi-modal Transformer (MMKET). Comprehensiveexperimental results show the effectiveness of the proposed model and provethat the debias operation has a great impact on the classification performancefor mERC. We hope our study will benefit the development of bias mitigation inmERC and related emotion studies.

A Multibias-mitigated and Sentiment Knowledge Enriched Transformer for Debiasing in Multimodal Conversational Emotion Recognition