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

Frequency domain processing, and in particular the use of Modified DiscreteCosine Transform (MDCT), is the most widespread approach to audio coding.However, at low bitrates, audio quality, especially for speech, degradesdrastically due to the lack of available bits to directly code the transformcoefficients. Traditionally, post-filtering has been used to mitigate artefactsin the coded speech by exploiting a-priori information of the source and extratransmitted parameters. Recently, data-driven post-filters have shown betterresults, but at the cost of significant additional complexity and delay. Inthis work, we propose a mask-based post-filter operating directly in MDCTdomain of the codec, inducing no extra delay. The real-valued mask is appliedto the quantized MDCT coefficients and is estimated from a relativelylightweight convolutional encoder-decoder network. Our solution is tested onthe recently standardized low-delay, low-complexity codec (LC3) at lowestpossible bitrate of 16 kbps. Objective and subjective assessments clearly showthe advantage of this approach over the conventional post-filter, with anaverage improvement of 10 MUSHRA points over the LC3 coded speech.

A DNN Based Post-Filter to Enhance the Quality of Coded Speech in MDCT Domain