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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Identifying multiple speakers without knowing where a speaker's voice is in arecording is a challenging task. In this paper, a hierarchical attentionnetwork is proposed to solve a weakly labelled speaker identification problem.The use of a hierarchical structure, consisting of a frame-level encoder and asegment-level encoder, aims to learn speaker related information locally andglobally. Speech streams are segmented into fragments. The frame-level encoderwith attention learns features and highlights the target related frameslocally, and output a fragment based embedding. The segment-level encoder workswith a second attention layer to emphasize the fragments probably related totarget speakers. The global information is finally collected from segment-levelmodule to predict speakers via a classifier. To evaluate the effectiveness ofthe proposed approach, artificial datasets based on Switchboard Cellular part1(SWBC) and Voxceleb1 are constructed in two conditions, where speakers' voicesare overlapped and not overlapped. Comparing to two baselines the obtainedresults show that the proposed approach can achieve better performances.Moreover, further experiments are conducted to evaluate the impact of utterancesegmentation. The results show that a reasonable segmentation can slightlyimprove identification performances.

Weakly Supervised Training of Hierarchical Attention Networks for Speaker Identification