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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Speech emotion recognition is a challenge and an important step towards morenatural human-computer interaction (HCI). The popular approach is multimodalemotion recognition based on model-level fusion, which means that themultimodal signals can be encoded to acquire embeddings, and then theembeddings are concatenated together for the final classification. However, dueto the influence of noise or other factors, each modality does not always tendto the same emotional category, which affects the generalization of a model. Inthis paper, we propose a novel regularization method via contrastive learningfor multimodal emotion recognition using audio and text. By introducing adiscriminator to distinguish the difference between the same and differentemotional pairs, we explicitly restrict the latent code of each modality tocontain the same emotional information, so as to reduce the noise interferenceand get more discriminative representation. Experiments are performed on thestandard IEMOCAP dataset for 4-class emotion recognition. The results show asignificant improvement of 1.44\% and 1.53\% in terms of weighted accuracy (WA)and unweighted accuracy (UA) compared to the baseline system.

Contrastive Regularization for Multimodal Emotion Recognition Using Audio and Text