The design of Acoustic-Phonetic decoders involves the identification of the acoustic units as a final step. In this paper two automatic classifiers for the Spanish unvoiced occlusives are presented. Only the acoustic features of the burst segment, automatically segmented from the speech waveform, were considered in the parameters estimation. The analysis of these features was carried out in both the time and frequency domains. In the first case, the classifier was designed in a procedural form. Alternatively, in the second case a statistical classifier was obtained from a previous discriminant analysis of the parameters. Both classifiers were tested over a CV context corpus uttered by 40 speakers not included in the training corpus, resulting in a good rate of identification.
Keywords: Speech recognition, Spanish occlusive features, acoustic-phonetic decoding, knowledge based.
Bibliographic reference. Torres, I. / Iparraguirre, P. (1993): "Acoustic-phonetic decoding of Spanish occlusive consonants", In EUROSPEECH'93, 457-460.