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

Self-supervised pre-trained speech models were shown effective for variousdownstream speech processing tasks. Since they are mainly pre-trained to mapinput speech to pseudo-labels, the resulting representations are only effectivefor the type of pre-train data used, either clean or mixture speech. With theidea of selective auditory attention, we propose a novel pre-training solutioncalled Selective-HuBERT, or SHuBERT, which learns the selective extraction oftarget speech representations from either clean or mixture speech.Specifically, SHuBERT is trained to predict pseudo labels of a target speaker,conditioned on an enrolled speech from the target speaker. By doing so, SHuBERTis expected to selectively attend to the target speaker in a complex acousticenvironment, thus benefiting various downstream tasks. We further introduce adual-path training strategy and use the cross-correlation constraint betweenthe two branches to encourage the model to generate noise-invariantrepresentation. Experiments on SUPERB benchmark and LibriMix datasetdemonstrate the universality and noise-robustness of SHuBERT. Furthermore, wefind that our high-quality representation can be easily integrated withconventional supervised learning methods to achieve significant performance,even under extremely low-resource labeled data.

Selective HuBERT: Self-Supervised Pre-Training for Target Speaker in Clean and Mixture Speech