Classification of lexical stress using spectral and prosodic features for computer-assisted language learning systems

• Autores: Ferrer, Luciana; Bratt, Harry; Richey, Colleen; Franco, Horacio; Abrash, Victor; Precoda, Kristin
• Año de publicación: 2015
• Idioma: inglés
• Tipo de recurso: artículo
• Materia: Computer-Assisted Language Learning; Gaussian Mixture Models; Lexical Stress Detection; Mel Frequency Cepstral Coefficients; Prosodic Features
• Institución: Consejo Nacional de Investigaciones Científicas y Técnicas

We present a system for detection of lexical stress in English words spoken by English learners. This system was designed to be part of the EduSpeak® computer-assisted language learning (CALL) software. The system uses both prosodic and spectral features to detect the level of stress (unstressed, primary or secondary) for each syllable in a word. Features are computed on the vowels and include normalized energy, pitch, spectral tilt, and duration measurements, as well as log-posterior probabilities obtained from the frame-level mel-frequency cepstral coefficients (MFCCs). Gaussian mixture models (GMMs) are used to represent the distribution of these features for each stress class. The system is trained on utterances by L1-English children and tested on English speech from L1-English children and L1-Japanese children with variable levels of English proficiency. Since it is trained on data from L1-English speakers, the system can be used on English utterances spoken by speakers of any L1 without retraining. Furthermore, automatically determined stress patterns are used as the intended target; therefore, hand-labeling of training data is not required. This allows us to use a large amount of data for training the system. Our algorithm results in an error rate of approximately 11% on English utterances from L1-English speakers and 20% on English utterances from L1-Japanese speakers. We show that all features, both spectral and prosodic, are necessary for achievement of optimal performance on the data from L1-English speakers; MFCC log-posterior probability features are the single best set of features, followed by duration, energy, pitch and finally, spectral tilt features. For English utterances from L1-Japanese speakers, energy, MFCC log-posterior probabilities and duration are the most important features.
Fil: Ferrer, Luciana. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Sri International; Estados Unidos
Fil: Bratt, Harry. Sri International; Estados Unidos
Fil: Richey, Colleen. Sri International; Estados Unidos
Fil: Franco, Horacio. Sri International; Estados Unidos
Fil: Abrash, Victor. Sri International; Estados Unidos
Fil: Precoda, Kristin. Sri International; Estados Unidos