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

Binaural speech separation in real-world scenarios often involves movingspeakers. Most current speech separation methods use utterance-levelpermutation invariant training (u-PIT) for training. In inference time,however, the order of outputs can be inconsistent over time particularly inlong-form speech separation. This situation which is referred to as the speakerswap problem is even more problematic when speakers constantly move in spaceand therefore poses a challenge for consistent placement of speakers in outputchannels. Here, we describe a real-time binaural speech separation model basedon a Wavesplit network to mitigate the speaker swap problem for moving speakerseparation. Our model computes a speaker embedding for each speaker at eachtime frame from the mixed audio, aggregates embeddings using online clustering,and uses cluster centroids as speaker profiles to track each speaker throughoutthe long duration. Experimental results on reverberant, long-form movingmultitalker speech separation show that the proposed method is less prone tospeaker swap and achieves comparable performance with u-PIT based models withground truth tracking in both separation accuracy and preserving the interauralcues.

Online Binaural Speech Separation of Moving Speakers With a Wavesplit Network