Burmese speech word recognition

• Main Author: Zaw, Wai Phyo
• Other Authors: Soon Ing Yann
• Format: Theses and Dissertations
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
• Online Access: http://hdl.handle.net/10356/41421
• Institution: Nanyang Technological University

The aim of this study is to develop a Burmese speech word recognition system. A phone level (sub-word) based voice-operated interface for phone dialing system is implemented using the Hidden Markov Model Tool Kit (HTK). The Mel Frequency Cepstral Coefficients (MFCCs) and Perceptual Linear Prediction Cepstral Coefficients (PLPCCs) are employed as feature in this project. In this study, the method called Half-Wave Rectification is applied to the MFCC to enhance the spectral peaks. The recognizer is evaluated with MFCCs and PLPCCs as well as rectified MFCCs. The evaluated results are discussed and finally, this study also provides the recommendation and future works to extend this project.