Comparison of parameter sets for temporal decomposition of speech

Temporal decomposition of a. speech utterance results in a description of speech panmeters in terms of overlapping target functions and associated target vectors. Although developed for economical speech coding, this method also provide! an interesting tool for deriving phonetic information from the acoustic speech signal. The target vectors may correspond to idealized articulatory targets; the target fun ctions describe the temporal evolution of these targets. The speech parameters used by Atal whe n he proposed this method (1983) are the log-area parameters. Our modified temporal decomposition method (1987) alro works with these parameters as input.. However, in principle, most commonly used parameter sets can be used. In t his paper we compare the results for six different sets of speech parameters . The main performance criterion will be the phonetic relevance of the target functions. The phonetic interpretation of the target vectors and the resynthesis of the speech signal will also be considered as criteria. From our experiments, we will conclude that the filter bank output para.meters and the log-area parameters are the most suitable parameter set! available for temporal decomposition.