A Modified Speech Recognition Algorithm for People with Physical Disabilities
Khalid Ahmad Darabkh, Laila Haddad, Saadeh Z. Sweidan · 2017
In this paper, we present a speech recognition algorithm that is based on double threshold voice activity detection (VAD) and Mel-frequency cepstral coefficients (MFCC). Interestingly, our proposed algorithm achieved a recognition rate of 98.7%. Our proposed approach was validated through conducting comparison with a relevant approach and its recognition rate was noticeably higher.