The use of LSF-based phonetic classification in low-rate coder design

In this paper we investigate a novel approach to low-bit rate quantisation of the spectral parameters of speech. This approach incorporates phonetic information into the structure of line spectral frequency (LSF) codebooks. As clear relationships exist between phonetic segments and LSFs, phonetic events can be expressed in terms of the structure of an LSF codebook and the successive vectors chosen by it. The investigation leads to the conclusion that the structure of LSF codebooks can be usefully employed in phonetic classification as a front end to multi-modal phonetic vocoding.