Syllable-based automatic Arabic speech recognition in different conditions of noise
Mohamed M. Azmi, Hesham Tolba · 2008
The presence of noise degrades the recognition percent of automatic speech recognition systems. The improvement of noise can be achieved by changing acoustic units during the recognition process. In this paper, we concentrate on automatic Arabic speech recognition in different conditions of noise using different acoustic units. Automatic Arabic speech was described by showing their constructing monophones, triphones and syllables. Speaker-independent hidden Markov models (HMMs)-based speech recognition system was designed using hidden Markov model toolkit (HTK). The database used of Arabic consists from fifty-nine Egyptian speakers. Speakers were asked to utter different sentences of Egyptian proverbs. As shown in experiments here, the recognition rates using syllables outperform monophones and triphones by 21.46% and 15.63% repectively, when SNR is 20 dB at the average of the some different conditions of noise. Motivated by the obtained results, speech recognition using syllables is more robustness to noise than triphones and monophones.