Modèle de Markov évolutif pour les tâches de suivi de locuteurs

Abstract : Seeking within a speech sequence the speaker utterances is one of the main tasks of indexing. In this paper, the proposed speaker tracking system is defined in the case where all speaker identities are known beforehand. The conversation is modeled as an evolutive Markov Model, in which speaker models computed are added one by one. A temporary indexing is proposed after each speaker adding and then challenged at the next step. This process is iterated until all the speakers are detected.The system has been assessed using multi-speaker messages generated by concatenation of Switchboard mono-speaker segments.The obtained results show the potentiality of the proposed solution.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01434692
Contributor : Sylvain Meignier <>
Submitted on : Wednesday, March 22, 2017 - 12:28:30 AM
Last modification on : Saturday, June 15, 2019 - 12:24:17 PM
Long-term archiving on : Friday, June 23, 2017 - 12:27:31 PM

File

jep2000.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01434692, version 1

Collections

Citation

Sylvain Meignier, Jean-François Bonastre, Corinne Fredouille, Teva Merlin. Modèle de Markov évolutif pour les tâches de suivi de locuteurs. JEP 2000, 2000, Aussois, France. pp.4. ⟨hal-01434692⟩

Share

Metrics

Record views

100

Files downloads

18