Abstract
An algorithm for the automatic generation of intonation, which had been developed within the framevork of the SPIN/ASSP program, was applied to a sample text. This text was presented to four expert listeners. Their comments inspired several comparisons between lhe intonation in the synthetic version and the intonation in a version of the text read by a human speaker.
The results showed that intonational choices in reading aloud appear to be guided by syntactic and phonological factors. In particular, the following results were obtained. With respect to declination, a distinction should be made between intonation phrases in sentence-initial, medial and final position. The size of sentence-intemal declination resets appeared to be related to syntactic structure.
| Original language | English |
|---|---|
| Title of host publication | Analysis and synthesis of speech : strategic research towards high-quality text-to-speech generation |
| Editors | V.J. Heuven, van, L.C.W. Pols |
| Place of Publication | Berlin, New York |
| Publisher | Mouton de Gruyter |
| Pages | 241-259 |
| ISBN (Print) | 3-11-013588-4 |
| Publication status | Published - 1993 |