Fuzzy Logic based Nam Speech Recognition for Tamil Syllables

Background/Objectives: In this paper, we propose the speech recognition of tamil words from dump mute persons with NAM sensor. Methods/Statistical analysis : The NAM sensor collects the speech signal of the tamil words frequently used in day-to-day life. These frequently used words are classified using fuzzy logic, based on syllables and the words are recognized. Findings: Until now, the NAM speech recognition was implemented only for vowels. From the 50 sample of speech sound, we attained 85% of accuracy in recognition. Application/Improvements: System helps in recognizing the tamil words from dump mute persons.