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

Vector Quantized Variational AutoEncoders (VQ-VAE) are a powerfulrepresentation learning framework that can discover discrete groups of featuresfrom a speech signal without supervision. Until now, the VQ-VAE architecturehas previously modeled individual types of speech features, such as only phonesor only F0. This paper introduces an important extension to VQ-VAE for learningF0-related suprasegmental information simultaneously along with traditionalphone features.The proposed framework uses two encoders such that the F0trajectory and speech waveform are both input to the system, therefore twoseparate codebooks are learned. We used a WaveRNN vocoder as the decodercomponent of VQ-VAE. Our speaker-independent VQ-VAE was trained with raw speechwaveforms from multi-speaker Japanese speech databases. Experimental resultsshow that the proposed extension reduces F0 distortion of reconstructed speechfor all unseen test speakers, and results in significantly higher preferencescores from a listening test. We additionally conducted experiments usingsingle-speaker Mandarin speech to demonstrate advantages of our architecture inanother language which relies heavily on F0.

Improved Prosody from Learned F0 Codebook Representations for VQ-VAE Speech Waveform Reconstruction