Generation and Analysis of English Phonetic Stress Modeling Dependent on Auto-context Algorithm

Stress is not only a component of the English phonetic structure but also a distinguisher of word meaning and part of speech. Moreover, it is the foundation of intonation and speech rhythm structure. In this paper, an auto-context algorithm conversion method is proposed based on the features of English focus stress expression. Data analysis suggests that the change in acoustic features of each syllable from neutral to focal phonetics is related to the position of the syllable relative to the focal point. Hence, the conversion model is applied to neutral phonetics, and a subjective evaluation test is performed on the conversion results. The experimental results suggest that the model established based on the auto-context algorithm has expressed the focus stress properly and achieved the purpose of improving the English phonetic expressiveness.