Please use this identifier to cite or link to this item:
Title: Articulatory Feature Extraction from Voice and Their Impact on Hybrid Acoustic Models
Authors: Llombart, Jorge
Miguel, Antonio
Lleida, Eduardo
UNESCO Clasification: 220990 Tratamiento digital. Imágenes
Keywords: Articulatory features
Neural network
Hybrid models
Issue Date: 2014
Publisher: Springer 
Journal: Lecture Notes in Computer Science 
Abstract: There is a great amount of information in the speech signal, although current speech recognizers do not exploit it completely. In this paper articulatory information is extracted from speech and fused to standard acoustic models to obtain a better hybrid acoustic model which provides improvements on speech recognition. The paper also studies the best input signal for the system in terms of type of speech features and time resolution to obtain a better articulatory information extractor. Then this information is fused to a standard acoustic model obtained with neural networks to perform the speech recognition achieving better results.
ISBN: 978-3-319-13622-6
ISSN: 0302-9743
DOI: 10.1007/978-3-319-13623-3_15
Source: Advances in Speech and Language Technologies for Iberian Languages. Lecture Notes in Computer Science, v. 8854 LNCS, p. 138-147
Appears in Collections:Capítulo de libro
Show full item record

Google ScholarTM




Export metadata

Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.