Deep belief networks for i-vector based speaker recognition

Inicio | ¿Qué es? | Contacto

English | Català

Consultar RECERCAT

Por comunidades y
colecciones Por fecha Por autores Por títulos Por temas (CDU)

Consultar departamento

Por fecha Por autores Por títulos Por temas (CDU)

Estadisticas

Del documento Todo RECERCAT

Mi RECERCAT

Entrar Alertas por correo-e

Directorio de otros repositorios

RECERCAT Principal > Universitat Politècnica de Catalunya > Documents de recerca > Visualizar documento

Para acceder a los documentos con el texto completo, por favor, siga el siguiente enlace: http://hdl.handle.net/2117/27145

dc.contributor	Universitat Politècnica de Catalunya. Departament de Teoria del Senyal i Comunicacions
dc.contributor	Universitat Politècnica de Catalunya. VEU - Grup de Tractament de la Parla
dc.contributor.author	Ghahabi Esfahani, Omid
dc.contributor.author	Hernando Pericás, Francisco Javier
dc.date	2014
dc.identifier.citation	Ghahabi, O.; Hernando, J. Deep belief networks for i-vector based speaker recognition. A: IEEE International Conference on Acoustics, Speech, and Signal Processing. "2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2014): Florence, Italy, 4-9 May 2014". Florència: Institute of Electrical and Electronics Engineers (IEEE), 2014, p. 1700-1704.
dc.identifier.citation	978-1-4799-2894-1
dc.identifier.citation	10.1109/ICASSP.2014.6853888
dc.identifier.uri	http://hdl.handle.net/2117/27145
dc.language.iso	eng
dc.publisher	Institute of Electrical and Electronics Engineers (IEEE)
dc.rights	info:eu-repo/semantics/openAccess
dc.subject	Àrees temàtiques de la UPC::Enginyeria de la telecomunicació
dc.subject	Àrees temàtiques de la UPC::Informàtica
dc.subject	Neural networks (Computer science)
dc.subject	Deep belief network
dc.subject	i-vector
dc.subject	Neural network
dc.subject	Speaker recognition
dc.subject	Xarxes neuronals (Informàtica)
dc.title	Deep belief networks for i-vector based speaker recognition
dc.type	info:eu-repo/semantics/publishedVersion
dc.type	info:eu-repo/semantics/conferenceObject
dc.description.abstract	The use of Deep Belief Networks (DBNs) is proposed in this paper to model discriminatively target and impostor i-vectors in a speaker verification task. The authors propose to adapt the network parameters of each speaker from a background model, which will be referred to as Universal DBN (UDBN). It is also suggested to backpropagate class errors up to only one layer for few iterations before to train the network. Additionally, an impostor selection method is introduced which helps the DBN to outperform the cosine distance classifier. The evaluation is performed on the core test condition of the NIST SRE 2006 corpora, and it is shown that 10% and 8% relative improvements of EER and minDCF can be achieved, respectively.
dc.description.abstract	Peer Reviewed

Mostrar el registro sencillo del ítem

Documentos relacionados

Otros documentos del mismo autor/a

Global impostor selection for DBNs in multi-session i-vector speaker recognition

Ghahabi Esfahani, Omid; Hernando Pericás, Francisco Javier

i-Vector modeling with deep belief networks for multi-session speaker recognition

Ghahabi Esfahani, Omid; Hernando Pericás, Francisco Javier

Deep neural networks for i-vector language identification of short utterances in cars

Ghahabi Esfahani, Omid; Bonafonte Cávez, Antonio; Hernando Pericás, Francisco Javier; Moreno Bilbao, M. Asunción

On the acoustic environment of a neonatal intensive care unit: initial description, and detection of equipment alarms

Raboshchuk, Ganna; Nadeu Camprubí, Climent; Ghahabi Esfahani, Omid; Solvez, Sergi; Muñoz Mahamud, Blanca; Riverola de Veciana, Ana; Navarro Hervas, Santiago

On the improvement of speaker diarization by detecting overlapped speech

Hernando Pericás, Francisco Javier; Hernando Pericás, Francisco Javier

Accesibilidad | Aviso legal | Política de Cookies | Documentos de uso interno

Coordinación

Patrocinio