Speech recognition using ANN and predator-influenced civilized swarm optimization algorithm

This paper proposes a hybrid optimization technique, predator-influenced civilized swarm optimization, by integrating civilized swarm optimization (CSO) and predator prey optimization (PPO) techniques. CSO is the integration of the attributes of particle swarm optimization and a society civilization algorithm (SCA). In the SCA, the swarm is divided into a few societies, and each society has its own society leader (SL); other individuals of the society are termed society members. The combination of all such societies forms a civilization, and the best-performing SL becomes the civilization leader (CL). In CSO, SLs and members update their positions through the guidance of their own CL and SLs, respectively, along with their best positions. In the proposed technique, the PPO technique is integrated with CSO, in which a predator particle is included in the swarm. The predator always tries to chase the CL in a controlled manner, which maintains diversity in the population and avoids local optimum solutions. The proposed optimization technique is applied to optimize the weights and biases of an artificial neural network (ANN) trained for speech recognition. Two databases have been used; one is a TI-46 isolated word database in clean and noisy conditions, and the other is a self-recorded Hindi numeral database. To evaluate the performance of the proposed technique, 2 performance criteria, correlation coefficient and mean square error, are applied. The results obtained by an ANN with the proposed technique outperform the results obtained by an ANN trained with particle swarm optimization, PPO, CSO, and backpropagation techniques in terms of correlation coefficient and mean square error.

PDF

___

[1] Kennedy J, Eberhart R. Particle swarm optimization. In: IEEE 1995 International Conference on Neural Networks; 27 November1 December 1995; Perth, Australia. Piscataway, NJ, USA: IEEE. pp. 1942-1948.
[2] Silva A, Neves A, Costa E. An empirical comparison of particle swarm and predator prey optimization. In: ONeill M, Sutcliffe RFE, Ryan C, Eaton M, Griffith NJL, editors. Artificial Intelligence and Cognitive Science. Berlin, Germany: Springer, 2002. pp. 103-110.
[3] Costa E, Silva A, Coelho LDS, Lebensztajn L. Multiobjective biogeography based optimization based on predator prey approach. IEEE T Magn 2012; 48: 951-954.
[4] Selvakumar AI, Thanushkodi K. Optimization using civilized swarm: solution to economic dispatch with multiple minima. Electr Power Syst Res 2009; 79: 8-16.
[5] Ray T, Liew KM. Society and civilization: an optimization algorithm based on the simulation of social behaviour. IEEE T Evol Comput 2003; 7: 386-396.
[6] Mehrotra K, Mohan CK, Ranka S. Elements of Artificial Neural Networks. Cambridge, MA, USA: MIT Press, 1997.
[7] Haykin S. Neural Networks: A Comprehensive Foundation. 1st ed. New York, NY, USA: Macmillan, 1991.
[8] Man ZH, Lee K, Wang DH, Cao ZW, Khoo SY. Robust singlehidden layer feedforward network-based pattern classifier. IEEE T Neural Networks Learn Syst 2012; 23: 1974-1986.
[9] Bilski J, Rutkowski L. A fast training algorithm for neural networks. IEEE T Circuit Syst Express Briefs 1998; 45: 1580-1591.
[10] Zhang R, Xu ZB, Huang GB, Wang DH. Global convergence of online BP training with dynamic learning rate. IEEE T Neural Networks Learn Syst 2012; 23: 330-341.
[11] Al-Batah MS, Isa NAM, Zamli KZ, Azizli KA. Modified recursive least squares algorithm to train the hybrid multilayered perceptron network. Appl Soft Comput 2010; 10: 236-244.
[12] Wang L, Zou F, Yang D, Chen D, Jiang Q. An improved teachinglearningbased optimization with neighborhood search for application of ANN. Neurocomputing 2014; 143: 231-247.
[13] J`anos DP. Calibrating artificial neural networks by global optimization. Expert Syst Appl 2012; 39: 25-32.
[14] Das G, Pattnaik PK, Padhy SK. Artificial neural network trained by particle swarm optimization for nonlinear channel equalization. Expert Syst Appl 2014; 41: 3491-3496.
[15] Rabiner L, Juang BH. Fundamentals of Speech Recognition. 1st ed. New York, NY, USA: Pearson Education, 1993.
[16] Krishnan VRV, Jayakumar A, Anto PB. Speech recognition of isolated Malayalam words using wavelet features and artificial neural network. In: IEEE 2008 International Symposium on Electronic Design, Test and Applications; 2325 January, 2008; Hong Kong. Piscataway, NJ, USA: IEEE. pp. 240-243.
[17] Wu J, Chan C. Isolated word recognition by neural network models with cross-correlation coefficients for speech dynamics. IEEE T Pattern Anal Mach Intell 1993; 15: 1174-1185.
[18] Zebulun RS, Perelmuter G, Velasco M, Pacheco MA. A comparison of different spectral analysis models for speech recognition using neural networks. In: IEEE 1997 39th Midwest Symposium on Circuits and Systems; 1821 August 1996; Ames, IA, USA. Piscataway, NJ, USA: IEEE. pp. 1428-1431.
[19] Dede G, Sazli MH. Speech recognition with artificial neural networks. Digital Signal Process 2009; 20: 763-768.
[20] Jay GW, Lawrence RR, Chin-Hui L, Goldman ER. Automatic recognition of keywords in unconstrained speech using hidden Markov models. IEEE T Acoust Speech Signal Process 1990; 38: 1870-1878.
[21] Yoshua B, Renato DM, Giovanni F, Ralf K. Global optimization of a neural networkhidden Markov model hybrid. IEEE T Neural Networks 1992; 3: 252-259.
[22] Maheshwari NU, Kabilan AP, Venkatesh R. Speech recognition system based on phonemes using neural networks. Int J Comput Sci Network Secur 2009; 9: 148-153.
[23] TI 46Word Speaker Dependent Isolated Word Corpus. NSIT Speech Disc 1991; 71.1.
[24] Farooq O, Datta S. Mel filter-like admissible wavelet structures for speech recognition. IEEE Signal Proc Let 2001; 8: 196-198.