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

The majority of recent state-of-the-art speaker verification architecturesadopt multi-scale processing and frequency-channel attention mechanisms.Convolutional layers of these models typically have a fixed kernel size, e.g.,3 or 5. In this study, we further contribute to this line of research utilisinga selective kernel attention (SKA) mechanism. The SKA mechanism allows eachconvolutional layer to adaptively select the kernel size in a data-drivenfashion. It is based on an attention mechanism which exploits both frequencyand channel domain. We first apply existing SKA module to our baseline. Then wepropose two SKA variants where the first variant is applied in front of theECAPA-TDNN model and the other is combined with the Res2net backbone block.Through extensive experiments, we demonstrate that our two proposed SKAvariants consistently improves the performance and are complementary whentested on three different evaluation protocols.

Frequency and Multi-Scale Selective Kernel Attention for Speaker Verification