Date of Publication :7th May 2016
Abstract: ASR (Automatic Speech Recognition) is most important for human computer interaction. In spite of tremendous advance in the ASR technique it is still far-away from human level performance. HMM (Hidden Markov model) based ASR system has made advancement but faced lots of limitation in practical problems. ANN (Artificial Neural Network) was used to replace HMM but the attempt was not that successful when trying to model high level dependency. Then attempt was made to combine this two to get the better performance. ANN-HMM hybrid model was introduced to combine the benefit of both and to increase the overall performance of ASR system. Large number of different hybrid architecture has proposed in the literature. Objective of this review paper is to focus on ANN-HMM. Most of the part of the paper is focused on describing the architecture of HMM, Neural Networkand how to use ANN for improvement of speech recognition.
Reference :
-
[1] Lawrence R. Rabiner, “A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition” proceedings of IEEE, Vol. 77, No. 2, February 1989.
[2] L. E. Baum and T. Petrie, "Statistical inference for probabilistic functions of finite state Markovchains," Ann. Math. Stat.vol. 37, pp. 1554-1563,1966.
[3] NelsonMorgon and Herve A. Bourlard, "Neural Networks for Statistical Recognition of Continuous Speech," Proceedings of the IEEE, vol. 83, no. 5, May 1995.
[4] Michael D. Richard and Richard P. Lippmann,"Neural Network Classifiers Estimate Bayesian a posteriori Probabilities," Neural Computation Volume 3 issue 4 1991.
[5] Xian Tang, "Hybrid Hidden Markov Model and Artificial NeuralNetwork for Automatic Speech Recognition," Pacific-Asia Conference on Circuits,Communications and System 2009.
[6] T. Robinson, “An application of recurrent nets to phone probability estimation,” IEEE Trans. Neural Networks, vol. 5, up.1994.
[7] T. Poggio and F. Girosi, “Networks for approximation and learning,” Proc. IEEE vol. 78, pp. 1481-1497, Sept. 1989. [8] D. E. Rumelhart, G. E. Hinton and R. J. Williams, “Learning internal representations by error propagation,” in Parallel Distributed Processing, Eds. Cambridge MA: MIT Press, 1986, vol. 1, pp. 318-362.
[9] Selina Chu, Shrikanth Narayanan C.-C. Jay Kuo, "Environmental Sound Recognition With Time– Frequency Audio Features," IEEE transaction on audio, speech and language processing, VOL. 17, NO. 6, August 2009.
[10] Mark Gales and Steve Young, “The Application of Hidden Markov Models in Speech Recognition” foundations and Trendsin Signal Processing Vol. 1, No. 3 (2007) 195–304
[11] L. E. Baum and J. A. Egon “An inequality with applications to statistical estimation for probabilistic functions of a Markov process and to a model for ecology” Bull. Amer. Meteorol. Soc. vol. 73, pp. 360- 363, 1967.
[12] L. E. Baum, T. Petrie, G. Soules, and N. Weiss, “A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains” Ann. Math. Stat.vol. 41, no. 1, pp. 164-171, 1970.
[13] S. E. Levinson, L. R. Rabiner, M. M. Sondhi, “Speaker Independent Isolated Digit Recognition Using Hidden Markov Models” ICASSP 83, BOSTON.