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

Generating natural speech with a diverse and smooth prosody pattern is achallenging task. Although random sampling with phone-level prosodydistribution has been investigated to generate different prosody patterns, thediversity of the generated speech is still very limited and far from what canbe achieved by humans. This is largely due to the use of uni-modaldistribution, such as single Gaussian, in the prior works of phone-levelprosody modelling. In this work, we propose a novel approach that modelsphone-level prosodies with a GMM-based mixture density network(MDN) and thenextend it for multi-speaker TTS using speaker adaptation transforms of Gaussianmeans and variances. Furthermore, we show that we can clone the prosodies froma reference speech by sampling prosodies from the Gaussian components thatproduce the reference prosodies. Our experiments on LJSpeech and LibriTTSdataset show that the proposed method with GMM-based MDN not only achievessignificantly better diversity than using a single Gaussian in bothsingle-speaker and multi-speaker TTS, but also provides better naturalness. Theprosody cloning experiments demonstrate that the prosody similarity of theproposed method with GMM-based MDN is comparable to recent proposedfine-grained VAE while the target speaker similarity is better.

Phone-Level Prosody Modelling with GMM-Based MDN for Diverse and Controllable Speech Synthesis