Document Type : Research Article

Author

• Dr. Sumedha Tuteja, Mr. Kenate Beyane , Dr. Rashmi Jain, Dr. G. Radha Krishna Murthy , Dr. Abdul Razak

Abstract

This study introduces a novel pattern recognition method. Hidden Markov Models are used to build the system (HMMs). We modify the Hidden Markov Models Toolkit (HTK) to address the pattern recognition problem in this paper. HTK was created for the purpose of voice recognition research. The mean of feature vectors is used to define patterns at first. By adding headers and displaying them in consecutive frames, such feature vectors are transformed to HTK format. A Windowing function is applied to each one. HTK then uses feature vectors for training and recognition tests.For the trials, we used 1600 randomly generated patterns from sixteen different consumer groups. The obtained findings demonstrate the efficacy of the suggested method.

Keywords

• Pattern classification
• Hidden Markov Models
• Hidden Markov Models Toolkit