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

Real-world complex acoustic environments especially the ones with a lowsignal-to-noise ratio (SNR) will bring tremendous challenges to a keywordspotting (KWS) system. Inspired by the recent advances of neural speechenhancement and context bias in speech recognition, we propose a robust audiocontext bias based DCCRN-KWS model to address this challenge. We form the wholearchitecture as a multi-task learning framework for both denosing and keywordspotting, where the DCCRN encoder is connected with the KWS model. Helped withthe denoising task, we further introduce an audio context bias module toleverage the real keyword samples and bias the network to better iscriminatekeywords in noisy conditions. Feature merge and complex context linear modulesare also introduced to strength such discrimination and to effectively leveragecontextual information respectively. Experiments on the internal challengingdataset and the HIMIYA public dataset show that our DCCRN-KWS system issuperior in performance, while ablation study demonstrates the good design ofthe whole model.

DCCRN-KWS: an audio bias based model for noise robust small-footprint keyword spotting