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

In speaker diarisation, speaker embedding extraction models often suffer fromthe mismatch between their training loss functions and the speaker clusteringmethod. In this paper, we propose the method of spectral clustering-awarelearning of embeddings (SCALE) to address the mismatch. Specifically, besidesan angular prototype cal (AP) loss, SCALE uses a novel affinity matrix losswhich directly minimises the error between the affinity matrix estimated fromspeaker embeddings and the reference. SCALE also includes p-percentilethresholding and Gaussian blur as two important hyper-parameters for spectralclustering in training. Experiments on the AMI dataset showed that speakerembeddings obtained with SCALE achieved over 50% relative speaker error ratereductions using oracle segmentation, and over 30% relative diarisation errorrate reductions using automatic segmentation when compared to a strong baselinewith the AP-loss-based speaker embeddings.

Spectral Clustering-aware Learning of Embeddings for Speaker Diarisation