期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2013
卷号：3
期号：6
出版社：S.S. Mishra
摘要：This paper describes the best method for automatic segmentation of Punjabi speech. Punjabi is one the most widely used language. Consequently, Punjabi is a syllabic language, so syllables are found to be the best units for the preparation of speech database. Further the automatic segmentation of speech is proved to be the better approach than the manual segmentation. The basic characteristics of the speech can be evaluated by the zero crossing rate and short term energy. Due to local energy fluctuation the STE functions are not reliable for continuous speech. Thus the approach used for the segmentation of speech is by using the negative derivative of Fourier transformations i.e. the Group Delay method. The peaks and valleys are better resolved when the signal is in minimum phase in case of group delay approach.
关键词：Short Term Energy; Zero crossing rate; Automatic speech segmentation; Group delay; Syllable units; Punjabi
