Correlation Features and a Linear Transform Specific Reproducing Kernel

Abstract. In this paper we introduce two ideas for phoneme classification: First, we derive the necessary steps to integrate linear transform into the computation of reproducing kernels. This concept not restricted to phoneme classification and can be applied in a wider range of research subjects. Second, in the context of support vector machine (SVM) classification, correlation features based on MFCC-vectors are proposed as a substitute for the common first and second derivatives, and the theory of the first part is applied to the new features. Additionally, an SVM structure in the spirit of phoneme states is introduced. Relative classification improvements of 40.67% compared to stacked MFCC features of equal dimension encourage further research in this direction. Key words: Correlation, Hilbert space, Reproducing kernel, Phoneme classification

• Type: Journal
• Year: 2010
• Where: TSD
• Authors: Andreas Beschorner, Dietrich Klakow
• Keywords: Phoneme Classification, Relative Classification Improvements, Signal Processing, Support Vector Machine, TSD 2010