Classification of Bangla Alphabets Phoneme based on Audio Features using MLPC & SVM

In several fields like natural language understanding, human-computer interaction, Bengali voice recognition can have a major impact. Acoustical Bangla Alphabet detection technique is a component of Bengali voice recognition. In the proposed system we have used different audio features of the processed data. Limited works have been done on Bangla Alphabets detection due to lack of data. Some works have been done on Bangla digits or only with Bangla vowels or few consonants but not with all of the alphabets. In this work, we have used 39 alphabets for classification. Multilayer Perceptron Classifier (MLPC) and support vector machine (SVM) are used to classify the data more accurately. The proposed system works on About 4095 data where 99.27% and 92.33% accuracy are achieved by the MLPC and SVM resectively.