Optimized linear discriminant analysis for extracting robust speech features

Linear discriminant analysis (LDA) is a feature selection method in speech recognition. LDA finds transformations that maximizes the between-class scatter and minimizes within-class scatter. This transformation can be obtained in a class-dependent or class independent manner. In this paper, we propose a method to use class-dependent LDA for speech recognition and MFCC extraction. In addition, we propose a multidimensional genetic algorithm to optimize class- dependent LDA transformation matrix for robust MFCC extraction. For this purpose, we first use logarithm of clean speech Mel filter bank energies (LMFE) of each class to define within-class scatter and between-class scatter . Next, we obtain class-dependent LDA transformation matrix using multidimensional genetic algorithm (MGA) and use this matrix in place of DCT in MFCC feature extraction. The experimental results show that proposed speech recognition and optimization methods using class-dependent LDA, achieves to a significant isolated word recognition rate on Aurora2 database.