DocumentCode :
1835595
Title :
Evaluating vowel pronunciation quality: Formant space matching versus ASR confidence scoring
Author :
Patil, Ashish ; Gupta, Chitralekha ; Rao, Preeti
Author_Institution :
Dept. of ECE, NIT Tiruchirappalli, Tiruchirappalli, India
fYear :
2010
fDate :
29-31 Jan. 2010
Firstpage :
1
Lastpage :
5
Abstract :
Quantitative evaluation of the quality of a speaker´s pronunciation of the vowels of a language can contribute to the important task of speaker accent detection. Our aim is to qualitatively and quantitatively distinguish between native and non-native speakers of a language on the basis of a comparative study of two analysis methods. One deals with relative positions of their vowels in formant (F1-F2) space that conveys important articulatory information. The other method exploits the sensitivity of trained phone models to accent variations, as captured by the log likelihood scores, to distinguish between native and non-native speakers.
Keywords :
maximum likelihood estimation; speech recognition; ASR confidence scoring; automatic speech recognition; formant space matching; log likelihood score; speaker accent detection; vowel pronunciation quality; Acoustic pulses; Acoustic signal detection; Automatic speech recognition; Databases; Hidden Markov models; Loudspeakers; Natural languages; Speech recognition; Testing; Training data; ASR confidence scoring; accent detection; formants; pronunciation;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Communications (NCC), 2010 National Conference on
Conference_Location :
Chennai
Print_ISBN :
978-1-4244-6383-1
Type :
conf
DOI :
10.1109/NCC.2010.5430187
Filename :
5430187
Link To Document :
بازگشت