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

In this work, we tackle a denoising and dereverberation problem with asingle-stage framework. Although denoising and dereverberation may beconsidered two separate challenging tasks, and thus, two modules are typicallyrequired for each task, we show that a single deep network can be shared tosolve the two problems. To this end, we propose a new masking method calledphase-aware beta-sigmoid mask (PHM), which reuses the estimated magnitudevalues to estimate the clean phase by respecting the triangle inequality in thecomplex domain between three signal components such as mixture, source and therest. Two PHMs are used to deal with direct and reverberant source, whichallows controlling the proportion of reverberation in the enhanced speech atinference time. In addition, to improve the speech enhancement performance, wepropose a new time-domain loss function and show a reasonable performance gaincompared to MSE loss in the complex domain. Finally, to achieve a real-timeinference, an optimization strategy for U-Net is proposed which significantlyreduces the computational overhead up to 88.9% compared to the na\"ive version.

Phase-aware Single-stage Speech Denoising and Dereverberation with U-Net