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

Automatic dubbing, which generates a corresponding version of the inputspeech in another language, could be widely utilized in many real-worldscenarios such as video and game localization. In addition to synthesizing thetranslated scripts, automatic dubbing needs to further transfer the speakingstyle in the original language to the dubbed speeches to give audiences theimpression that the characters are speaking in their native tongue. However,state-of-the-art automatic dubbing systems only model the transfer on durationand speaking rate, neglecting the other aspects in speaking style such asemotion, intonation and emphasis which are also crucial to fully perform thecharacters and speech understanding. In this paper, we propose a jointmulti-scale cross-lingual speaking style transfer framework to simultaneouslymodel the bidirectional speaking style transfer between languages at bothglobal (i.e. utterance level) and local (i.e. word level) scales. The globaland local speaking styles in each language are extracted and utilized topredicted the global and local speaking styles in the other language with anencoder-decoder framework for each direction and a shared bidirectionalattention mechanism for both directions. A multi-scale speaking style enhancedFastSpeech 2 is then utilized to synthesize the predicted the global and localspeaking styles to speech for each language. Experiment results demonstrate theeffectiveness of our proposed framework, which outperforms a baseline with onlyduration transfer in both objective and subjective evaluations.

Joint Multi-scale Cross-lingual Speaking Style Transfer with Bidirectional Attention Mechanism for Automatic Dubbing