Lexical stress detection on stress-minimal word pairs

We present a study on the use of lexical stress classi cation to aid in the recognition of phonetically similar words. In this study, we use a simple pattern recognition approach to determine which syllable is lexically stressed for phonetically similar word pairs (e.g., PERfect, perFECT) extracted from continuously spoken sentences. We use a combination of two features from the acoustic correlates of lexical stress, and assume multivariate Gaussian distributions to form a Bayesian classi er. The features used are normalized energy and duration of the vowel for each syllable of the word. We evaluate several normalization methods. Two sets of sentences were designed for this study. For the pilot experiment, the classi cation accuracy on words from the natural sentence set was 89.9% and on words from the control sentence set was 100%. To improve the performance, three-feature classi ers, which included two normalized energy features and one normalized duration feature, were developed. The classi cation accuracy on words from the natural sentence set was 97.23%.