Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
| Paper: | SP-P9.10 |
| Session: | Topics in Speech Synthesis |
| Time: | Wednesday, May 19, 15:30 - 17:30 |
| Presentation: |
Poster |
| Topic: |
Speech Processing: Speech Synthesis (including TTS) |
| Title: |
A STRATEGY TO SOLVE DATA SCARCITY PROBLEMS IN CORPUS BASED INTONATION MODELLING |
| Authors: |
Valentín Cardeñoso; University of Valladolid | | |
| | David Escudero; University of Valladolid | | |
| Abstract: |
The problem of data scarcity in corpus-based intonation modelling for TTS applications is addressed in this communication. Use of multiple model dictionaries is proposed to predict patterns unseen in the training corpus.To improve the models in classes without enough number of samples, a grouping strategy is proposed and an experimental study of it is presented. Predicted pitch profiles closer to the real ones can be obtained by this means. |
| |
| Back | |