Skip to Main content Skip to Navigation
Conference papers

Speaker diarization: about whom the speaker is talking?

Abstract : The automatic speaker diarization consists in splitting the signal into homogeneous segments and clustering them by speakers. However the speaker segments are specified with anonymous labels. This pa- per proposed a solution to identify those speakers by extracting their full names pronounced in the show. With a semantic classification tree automatically built on a training corpus, the full names detected in transcription of a segment are associated to this segment or to one of its neighbors. Then, a merging method allows to associate a full name to a speaker cluster instead of a anonymous label provided by the diarization. The experiments are carried out over French broadcast news records from the ESTER 2005 evaluation campaign. About 70% show duration is correctly processed for both development and eval- uation corpora. On the evaluation corpus, 18.15% show duration is wrongly named and no decision is taken for 11.91% show duration.
Complete list of metadatas

Cited literature [13 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01434121
Contributor : Sylvain Meignier <>
Submitted on : Thursday, February 9, 2017 - 2:14:03 PM
Last modification on : Thursday, December 21, 2017 - 12:56:46 AM
Document(s) archivé(s) le : Wednesday, May 10, 2017 - 1:52:53 PM

File

odyssey.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01434121, version 1

Collections

Citation

Julie Mauclair, Sylvain Meignier, Yannick Estève. Speaker diarization: about whom the speaker is talking?. IEEE Speaker Odyssey 2006, 2006, San Juan Puerto Rico. ⟨hal-01434121⟩

Share

Metrics

Record views

240

Files downloads

367