On the use of phone-gram units in recurrent neural networks for language identification

Christian Salamea, Luis Fernando D'Haro, Ricardo De Córdoba, Rubén San-Segundo

Resultado de la investigación: Contribución a una conferenciaDocumentorevisión exhaustiva

Resumen

In this paper we present our results on using RNN-based LM scores trained on different phone-gram orders and using different phonetic ASR recognizers. In order to avoid data sparseness problems and to reduce the vocabulary of all possible n-gram combinations, a K-means clustering procedure was performed using phone-vector embeddings as a pre-processing step. Additional experiments to optimize the amount of classes, batch-size, hidden neurons, state-unfolding, are also presented. We have worked with the KALAKA-3 database for the plenty-closed condition [1]. Thanks to our clustering technique and the combination of high level phone-grams, our phonotactic system performs ~13% better than the unigram-based RNNLM system. Also, the obtained RNNLM scores are calibrated and fused with other scores from an acoustic-based i-vector system and a traditional PPRLM system. This fusion provides additional improvements showing that they provide complementary information to the LID system.

Idioma originalInglés
Páginas117-123
Número de páginas7
DOI
EstadoPublicada - 1 ene 2016
EventoSpeaker and Language Recognition Workshop, Odyssey 2016 - Bilbao, Espana
Duración: 21 jun 201624 jun 2016

Conferencia

ConferenciaSpeaker and Language Recognition Workshop, Odyssey 2016
País/TerritorioEspana
CiudadBilbao
Período21/06/1624/06/16

Huella

Profundice en los temas de investigación de 'On the use of phone-gram units in recurrent neural networks for language identification'. En conjunto forman una huella única.

Citar esto