Speech Recognition with Emphasis on Wavelet based Feature Extraction

Abstract

In this paper some of the commonly used feature extraction techniques are presented and a new set of features based on the Discrete Wavelet Transform (DWT) and Admissible Wavelet Packet Transform (AWPT) is presented for the recognition of phonemes. These features overcome the problem of shift variance and speaker dependence encountered in the earlier features derived by using wavelet transform. Further study on the earlier proposed energy features derived by DWT is carried out and AWPT is proposed for phoneme recognition to overcome the problems with DWT based features. Further a new set of features based on the logarithmic compression of the energy is proposed which shows considerable improvement in the recognition performance.

Indexing terms:

• Phoneme recognition
• Feature extraction
• Wavelet transform
• Linear discriminant analysis

Additional information

Notes on contributors

O Farooq

O Farooq, obtained BSc Engineering and MSc Engineering degrees form Z H College of Engineering and Technology AMU in 1991 and 1993 respectively. He joined the department of Electronics Engineering as a lecturer in 1992 and is currently doing PhD at Loughborough University, UK under the Commonwealth scholarship. His area of research is phoneme recognition. He has authored and co-authored over 18 papers in refereed academic journals and national/international conference proceedings. He is a member of IETE, India and life member of ISTE India.

S Datta

S Datta, received BSc degree from the University of Calcutta and MSc and PhD degrees in Computer science from the University of London. He spent twenty years in industrial research related to information technology and advanced signal processing at the research and Advanced Development Centre, International Computers Ltd, where he worked as a Senior Research Consultant. Since joining Loughborough University in 1987 he has continued to work on speech processing and extended his research activities to include cursive script recognition and bioacoustics. He has authored and co-authored over 90 papers in refereed academic journals and international conference proceedings, and over 35 articles and reports in other publication categories, including editorship of a book and workshop proceedings. He is a fellow of IETE, India, member of IEE, UK, British Machine Vision Association, the Society of the Study of Artificial Intelligence and Simulation of Behaviour and the British Computer Society.