DocumentCode :
3585074
Title :
Classification of lexical stress patterns using deep neural network architecture
Author :
Shahin, Mostafa Ali ; Ahmed, Beena ; Ballard, Kirrie J.
Author_Institution :
Electr. & Comput. Eng. Program, Texas A&M Univ. at Qatar, Doha, Qatar
fYear :
2014
Firstpage :
478
Lastpage :
482
Abstract :
Lexical stress is a key diagnostic marker of disordered speech as it strongly affects speech perception. In this paper we introduce an automated method to classify between the different lexical stress patterns in children´s speech. A deep neural network is used to classify between strong-weak (SW), weak-strong (WS) and equal-stress (SS/WW) patterns in English by measuring the articulation change between the two successive syllables. The deep neural network architecture is trained using a set of acoustic features derived from pitch, duration and intensity measurements along with the energies in different frequency bands. We compared the performance of the deep neural classifier to a traditional single hidden layer MLP. Results show that the deep neural classifier outperforms the traditional MLP. The accuracy of the deep neural system is approximately 85% when classifying between the unequal stress patterns (SW/WS) and greater than 70% when classifying both equal and unequal stress patterns.
Keywords :
multilayer perceptrons; neural net architecture; pattern classification; speech processing; SS-WW patterns; SW patterns; WS patterns; articulation change; deep neural classifier; deep neural network architecture; disordered speech; equal-stress patterns; lexical stress pattern classification; single hidden layer MLP; speech perception; strong-weak patterns; unequal stress patterns; weak-strong patterns; Accuracy; Acoustics; Feature extraction; Neural networks; Speech; Stress; Training; automatic assessment; deep neural network; lexical stress; prosody;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Spoken Language Technology Workshop (SLT), 2014 IEEE
Type :
conf
DOI :
10.1109/SLT.2014.7078621
Filename :
7078621
Link To Document :
بازگشت