Segmentation of speech into syllable units using fuzzy smoothed short term energy contour

This paper investigates the problem of segmenting speech into sub_word units. In this work, a novel technique based on fuzzy smoothing is applied on short term energy function of speech wave. The smoothed energy contour is searched then in order to find local minima that imply to syllable units. Experimental results-conducted on 200 different sentences, 20 speakers (12 males, 8 females) in Farsdat¹ dataset-showed that about 93.8% of syllables are detected with less than 20% of average syllable duration error. The proposed method has a good improvement compared with common moving average smoothing method.