Abstract
This paper implemented an artificial neural network (ANN) on a field programmable gate array (FPGA) chip for Mandarin speech measurement and recognition of nonspecific speaker. A three-layer hybrid learning algorithm (HLA), which combines genetic algorithm (GA) and steepest descent method, was proposed to fulfill a faster global search of optimal weights in ANN. Some other popular evolutionary algorithms, such as differential evolution, particle swarm optimization and improve GA, were compared to the proposed HLA. It can be seen that the proposed HLA algorithm outperforms the other algorithms. Finally, the designed system was implemented on an FPGA chip with an SOC architecture to measure and recognize the speech signals.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Sivaram GSVS, Nemala SK, Mesgarani N, Hermansky H (2010) Data-driven and feedback based spectro-temporal features for speech recognition. IEEE Signal Process Lett 17(11):957–960
Lauria S (2007) Talking to machines: introducing Robot perception to resolve speech recognition uncertainties. Circuits Syst Signal Process 26(4):513–526
Wan CY, Lee LS (2008) Histogram-based quantization for robust and/or distributed speech recognition. IEEE Trans Audio Speech Lang Processing 16(4):859–873
Hagon MT, Demuth HB, Beale M (1996) Neural network design. Thomson Learning, Stamford
Kwong S, Chau CW (1997) Analysis of parallel genetic algorithms on HMM based speech recognition system. IEEE Trans Consumer Electron 43(4):1229–1233
Shi Y, Liu J, Liu R (2001) Single-chip speech recognition system based on 8051 microcontroller core. IEEE Trans Consumer Electron 47(1):149–153
Lin FJ, Huang PK, Chou WD (2007) Recurrent-fuzzy-neural-network-controlled linear induction motor servo drive using genetic algorithms. IEEE Trans Ind Electron 54(3):1449–1461
Karamalis PD, Kanatas AG, Constantinou P (2009) A genetic algorithm applied for optimization of antenna arrays used in mobile radio channel characterization devices. IEEE Trans Instrum Meas 58:2475–2487
Chu WC (2003) Speech coding algorithms. Wiley, Wiley-IEEE, New Jersey
Huang X, Acero A, Wuenon H (2005) Spoken language processing a guide to theory algorithm and system development. Pearson, London
Leung HF, Lam HK, Ling SH (2003) Tuning of the structure and parameters of a neural network using an improved genetic algorithm. IEEE Trans Neural Netw 14:79–88
Kennedy J, Eberhart RC (1995) “Particle swarm optimization.” In: Proceedings IEEE International Conference on Neural Networks (Perth, Australia), IEEE Service Center, Piscataway, NJ, pp IV:1942–1948, 1995
Storn R, Price K (1997) Differential evolution- A simple and efficient heurist for global optimization over continuous spaces. J. of Global Optimization 11:341–359
Runstein F, Violaro F (1995)”An isolated-word speech recognition system using neural networks”.In: Proceeding of the 38th midwest symposium on circuit and systems, Vol 1, pp 550–553, 1995
Sadaoki F, Dekker M (2001) Digital speech processing, synthesis, and recognition. Marcel Dekker, New York
Acknowledgments
This research work was supported by the National Science Council of the Republic of China under contract NSC 100-2221-E-390-025-MY2.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Pan, ST., Lan, ML. An efficient hybrid learning algorithm for neural network–based speech recognition systems on FPGA chip. Neural Comput & Applic 24, 1879–1885 (2014). https://doi.org/10.1007/s00521-013-1428-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-013-1428-5