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

Text to speech (TTS) has been broadly used to synthesize natural andintelligible speech in different scenarios. Deploying TTS in various enddevices such as mobile phones or embedded devices requires extremely smallmemory usage and inference latency. While non-autoregressive TTS models such asFastSpeech have achieved significantly faster inference speed thanautoregressive models, their model size and inference latency are still largefor the deployment in resource constrained devices. In this paper, we proposeLightSpeech, which leverages neural architecture search~(NAS) to automaticallydesign more lightweight and efficient models based on FastSpeech. We firstprofile the components of current FastSpeech model and carefully design a novelsearch space containing various lightweight and potentially effectivearchitectures. Then NAS is utilized to automatically discover well performingarchitectures within the search space. Experiments show that the modeldiscovered by our method achieves 15x model compression ratio and 6.5xinference speedup on CPU with on par voice quality. Audio demos are provided athttps://speechresearch.github.io/lightspeech.

LightSpeech: Lightweight and Fast Text to Speech with Neural Architecture Search