Sensor based speech production system without use of glottis

Abstract

The present work proposes a speech production mechanism using source-filter (Vocal tract) model, with the help of sensors placed in the oral cavity to synthesize speech. The speech is produced without involvement of glottis. This is achieved by exciting the articulatory system (Oral cavity) with sensor based input. The proposed system creates a virtual voice for a person who lost his voice due to some disability in vocal tract. The Vocal tract model and articulatory system has been tested with English alphabets (Vowels and consonants) using LabView, which has been validated by perceptual test. The present work mainly focuses on the role of the tongue as speech articulator.