The present paper addresses the question of the efficiency of independent Component Analysis (ICA) as a statistical process for deriving optimal representational bases for the projection of spectrum and cepstrum in the context of Automatic Speech Recognition (ASR). Several decorrelation strategies have been applied on the log-spectrum and cepstrum to fulfill the practical need of a diagonal covariance HMM for uncorrelated features. In our work we question the optimality of a fixed decorrelation strategy as dCT and follow an emerging trend in ASR that designs projection bases based on the statistics of speech. We differentiate oru approach from the second order statistics of Discrete Cosine Transform (DCT), Linear Discrimination Analyis (LDA) and Principal Component Analysis (PCA) by proposing an alternative data-driven approach based on HIgher Order Statistics.
展开▼