Genetic Programming for Automatic Stress Detection in Spoken English
This paper describes an approach to the use of genetic programming (GP) for the automatic detection of rhythmic stress in spoken New Zealand English. A linear-structured GP system uses speaker independent prosodic features and vowel quality features as terminals to classify each vowel segment as stressed or unstressed. Error rate is used as the fitness function. In addition to the standard four arithmetic operators, this approach also uses several other arithmetic, trigonometric, and conditional functions in the function set. The approach is evaluated on 60 female adult utterances with 703 vowels and a maximum accuracy of 92.61% is achieved. The approach is compared with decision trees (DT) and support vector machines (SVM). The results suggest that, on our data set, GP outperforms DT and SVM for stress detection, and GP has stronger automatic feature selection capability than DT and SVM.
KeywordsSupport Vector Machine Prosodic Feature Lexical Stress Vowel Quality Genetic Programming System
Unable to display preview. Download preview PDF.
- 1.Ladefoged, P.: Three Areas of experimental phonetics. Oxford University Press, London (1967)Google Scholar
- 2.Ladefoged, P.: A Course in Phonetics, 3rd edn. Harcourt Brace Jovanovich, New York (1993)Google Scholar
- 3.Waibel, A.: Recognition of lexical stress in a continuous speech system - a pattern recognition approach. In: Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, Tokyo, Japan, pp. 2287–2290 (1986)Google Scholar
- 4.Jenkin, K.L., Scordilis, M.S.: Development and comparison of three syllable stress classifiers. In: Proceedings of the International Conference on Spoken Language Processing, Philadelphia, USA, pp. 733–736 (1996)Google Scholar
- 5.van Kuijk, D., Boves, L.: Acoustic characteristics of lexical stress in continuous speech. In: Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, Munich, Germany, vol. 3, pp. 1655–1658 (1999)Google Scholar
- 6.Xie, H., Andreae, P., Zhang, M., Warren, P.: Detecting stress in spoken English using decision trees and support vector machines. Australian Computer Science Communications (Data Mining, CRPIT 32) 26, 145–150 (2004)Google Scholar
- 7.Conrads, M., Nordin, P., Banzhaf, W.: Speech sound discrimination with genetic programming. In: Proceedings of the First European Workshop on Genetic Programming, pp. 113–129 (1998)Google Scholar
- 8.Francone, F.D.: Discipulus owner’s manual (2004)Google Scholar
- 9.Xie, H., Andreae, P., Zhang, M., Warren, P.: Learning models for English speech recognition. Australian Computer Science Communications (Computer Science, CRPIT 26) 26, 323–330 (2004)Google Scholar
- 10.Quinlan, J.: C4.5: Programs for machine learning. Morgan Kaufmann, San Francisco (1993)Google Scholar
- 11.Chang, C.C., Lin, C.J.: LIBSVM: a library for support vector machines (2003), http://www.csie.ntu.edu.tw/_cjlin/papers/libsvm.pdf
- 13.Dy, J.G., Brodley, C.E.: Feature selection for unsupervised learning. Journal of Machine Learning Research 5, 845–889 (2004)Google Scholar