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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Deep learning-based speech enhancement has shown unprecedented performance inrecent years. The most popular mono speech enhancement frameworks areend-to-end networks mapping the noisy mixture into an estimate of the cleanspeech. With growing computational power and availability of multichannelmicrophone recordings, prior works have aimed to incorporate spatial statisticsalong with spectral information to boost up performance. Despite an improvementin enhancement performance of mono output, the spatial image preservation andsubjective evaluations have not gained much attention in the literature. Thispaper proposes a novel stereo-aware framework for speech enhancement, i.e., atraining loss for deep learning-based speech enhancement to preserve thespatial image while enhancing the stereo mixture. The proposed framework ismodel independent, hence it can be applied to any deep learning basedarchitecture. We provide an extensive objective and subjective evaluation ofthe trained models through a listening test. We show that by regularizing foran image preservation loss, the overall performance is improved, and the stereoaspect of the speech is better preserved.

A Training Framework for Stereo-Aware Speech Enhancement using Deep Neural Networks