ASIC design of a phoneme recogniser based on discrete wavelet transforms and support vector machines

This paper presents the design of an ASIC for the task of multi-speaker phoneme recognition in continuous speech environments. The phoneme recogniser is based on DWTs for feature extraction and the One-against-one SVM method, along a priorities scheme, for classification. The ASIC design was fabricated on an AMS 0.35μ CMOS C35B4C3 chip. The final ASIC design resulted into a chip size equal to 43.35mm2, with the requirement of an external memory storage of size 18.25Mb. Moreover, the ASIC design of the phoneme recogniser is approximately 4 times faster than the equivalent software-based approach and consumes 12.5mW, making it appealing to mobile devices. The performance results obtained from the ASIC design confirmed that this system is a promising basis for future hardware ASR systems.

[1]  Linda Dailey Paulson.  Speech Recognition Moves from Software to Hardware , 2006, Computer.

[2]  Ivan Grech,et al.  Hardware-based support vector machine for phoneme classification , 2013, Eurocon 2013.

[3]  Shivesh Ranjan.  A Discrete Wavelet Transform Based Approach to Hindi Speech Recognition , 2010, 2010 International Conference on Signal Acquisition and Processing.

[4]  Ivan Grech,et al.  Design of a hardware-based discrete wavelet transform architecture for phoneme recognition , 2014, 2014 6th International Symposium on Communications, Control and Signal Processing (ISCCSP).

[5]  Chung-Hsien Wu,et al.  VLSI Architecture and Implementation for Speech Recognizer Based on Discriminative Bayesian Neural Network , 2002, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[6]  Ivan Grech,et al.  Comparison of different multiclass SVM methods for speaker independent phoneme recognition , 2012, 2012 5th International Symposium on Communications, Control and Signal Processing.

[7]  Ivan Grech,et al.  Discrete wavelet transforms with multiclass SVM for phoneme recognition , 2013, Eurocon 2013.

[8]  Ivan Grech,et al.  A study on pitch variation on the use of DWT with SVM for speaker independent phoneme recognition , 2012, 2012 5th International Symposium on Communications, Control and Signal Processing.

[9]  Peng Li,et al.  Design of a Low-Power Coprocessor for Mid-Size Vocabulary Speech Recognition Systems , 2011, IEEE Transactions on Circuits and Systems I: Regular Papers.

[10]  Sylvio Barbon Junior,et al.  Improved Dynamic Time Warping Based on the Discrete Wavelet Transform , 2007, Ninth IEEE International Symposium on Multimedia Workshops (ISMW 2007).

[11]  Ivan Grech,et al.  Support Vector Machines with the priorities method for speaker independent phoneme recognition , 2011, 2011 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT).