Speech recognition using Adaboost HMM

• Main Author: Ooi, Mun Siang
• Other Authors: Foo Say Wei
• Format: Theses and Dissertations
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
• Online Access: http://hdl.handle.net/10356/64891
• Institution: Nanyang Technological University

For speech recognition, Hidden Markov Model (HMM) is a popular approach as the classifier with high degree of accuracy; Adaptive Boosting (Adaboost) is a method to improve the performance of a given base classifier. In this study, Adaboost technique is applied to HMM classifier in speech recognition to test the resulting performance. Experiments on several speech corpora showed that Adaboost-HMM classifiers are significantly more accurate than the baseline HMM classifiers. Results also showed that sufficient training samples that cover most of the entire sample space is necessary for generalization of Adaboost-HMM classifiers.