Speech Enhancement via Smart Larynx of Variable Frequency for Laryngectomee Patient for Tamil Language Syllables Using RADWT Algorithm

Abstract

The Laryngectomee patient has damage in vocal cord and leads to loss of speech; therefore, patient uses electrolarynx (EL) device during speech for speech clarity. The EL device consists of a vibrator placed over the vocal surface during speech, and the speech signal modulates with the vibrator signal for clarity in speech. However, vibration from the EL device produces a constant signal and produces less speech clarity during various environmental conditions such as indoor, outdoor and populated area. In this paper, the above problem is alleviated by generating automated vibration signal with manual tuning for speech clarity in different environmental conditions. The proposed Smart Larynx (SL) device replaces the electrolarynx (EL) which consists of a Smart Phone and Vibrator App with frequency ranging from 250 to 450 Hz. From the experimental result, the speech attains about 85% clarity of normal speech after processing with Radial Dilation Wavelet Transform algorithm. The SL device validates the clarity of the speech based on monosyllable, bisyllable and trisyllable in Tamil Language.