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

In recent years, speech emotion recognition (SER) has been used in wideranging applications, from healthcare to the commercial sector. In addition tosignal processing approaches, methods for SER now also use deep learningtechniques which provide transfer learning possibilities. However, generalizingover languages, corpora and recording conditions is still an open challenge. Inthis work we address this gap by exploring loss functions that aid intransferability, specifically to non-tonal languages. We propose a variationalautoencoder (VAE) with KL annealing and a semi-supervised VAE to obtain moreconsistent latent embedding distributions across data sets. To ensuretransferability, the distribution of the latent embedding should be similaracross non-tonal languages (data sets). We start by presenting a low-complexitySER based on a denoising-autoencoder, which achieves an unweightedclassification accuracy of over 52.09% for four-class emotion classification.This performance is comparable to that of similar baseline methods. Followingthis, we employ a VAE, the semi-supervised VAE and the VAE with KL annealing toobtain a more regularized latent space. We show that while the DAE has thehighest classification accuracy among the methods, the semi-supervised VAE hasa comparable classification accuracy and a more consistent latent embeddingdistribution over data sets.

Towards Transferable Speech Emotion Representation: On loss functions for cross-lingual latent representations