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

Identifying multiple speakers without knowing where a speaker's voice is in arecording is a challenging task. This paper proposes a hierarchical networkwith transformer encoders and memory mechanism to address this problem. Theproposed model contains a frame-level encoder and segment-level encoder, bothof them make use of the transformer encoder block. The multi-head attentionmechanism in the transformer structure could better capture different speakerproperties when the input utterance contains multiple speakers. The memorymechanism used in the frame-level encoders can build a recurrent connectionthat better capture long-term speaker features. The experiments are conductedon artificial datasets based on the Switchboard Cellular part1 (SWBC) andVoxceleb1 datasets. In different data construction scenarios (Concat andOverlap), the proposed model shows better performance comparaing with fourstrong baselines, reaching 13.3% and 10.5% relative improvement compared withH-vectors and S-vectors. The use of memory mechanism could reach 10.6% and 7.7%relative improvement compared with not using memory mechanism.

T-vectors: Weakly Supervised Speaker Identification Using Hierarchical Transformer Model