Fusion of MFCC & LPC Feature Sets for Accurate Speaker Identification
Pages : 1763-1766
Several feature extraction techniques are proposed for SRS. All these techniques achieve good recognition rate using ANN. To achieve good recognition rate with best performance is the objective of this paper. Using the two feature extraction techniques, features can concatenate & this technique will achieve 100% recognition rate with best performance using SCG algorithm. For this purpose MFCC & LPC techniques are used for extracting features.
Keywords: SRS- Speaker Recognition System, ANN- Artificial Neural Network, SCG-Scaled conjugate gradient, MFCC- Mel Frequency Cepstral Coefficients, LPC- Linear Predictive Coding.
Article published in International Journal of Current Engineering and Technology, Vol.3,No.5(Dec- 2013)