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

Denoising diffusion probabilistic models (DDPMs) have shown promisingperformance for speech synthesis. However, a large number of iterative stepsare required to achieve high sample quality, which restricts the inferencespeed. Maintaining sample quality while increasing sampling speed has become achallenging task. In this paper, we propose a "Co"nsistency "Mo"del-based"Speech" synthesis method, CoMoSpeech, which achieve speech synthesis through asingle diffusion sampling step while achieving high audio quality. Theconsistency constraint is applied to distill a consistency model from awell-designed diffusion-based teacher model, which ultimately yields superiorperformances in the distilled CoMoSpeech. Our experiments show that bygenerating audio recordings by a single sampling step, the CoMoSpeech achievesan inference speed more than 150 times faster than real-time on a single NVIDIAA100 GPU, which is comparable to FastSpeech2, making diffusion-sampling basedspeech synthesis truly practical. Meanwhile, objective and subjectiveevaluations on text-to-speech and singing voice synthesis show that theproposed teacher models yield the best audio quality, and the one-step samplingbased CoMoSpeech achieves the best inference speed with better or comparableaudio quality to other conventional multi-step diffusion model baselines. Audiosamples are available at https://comospeech.github.io/.

CoMoSpeech: One-Step Speech and Singing Voice Synthesis via Consistency Model