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

One-shot voice conversion (VC), which performs conversion across arbitraryspeakers with only a single target-speaker utterance for reference, can beeffectively achieved by speech representation disentanglement. Existing workgenerally ignores the correlation between different speech representationsduring training, which causes leakage of content information into the speakerrepresentation and thus degrades VC performance. To alleviate this issue, weemploy vector quantization (VQ) for content encoding and introduce mutualinformation (MI) as the correlation metric during training, to achieve properdisentanglement of content, speaker and pitch representations, by reducingtheir inter-dependencies in an unsupervised manner. Experimental resultsreflect the superiority of the proposed method in learning effectivedisentangled speech representations for retaining source linguistic content andintonation variations, while capturing target speaker characteristics. In doingso, the proposed approach achieves higher speech naturalness and speakersimilarity than current state-of-the-art one-shot VC systems. Our code,pre-trained models and demo are available athttps://github.com/Wendison/VQMIVC.

VQMIVC: Vector Quantization and Mutual Information-Based Unsupervised Speech Representation Disentanglement for One-shot Voice Conversion