Summary of the paper
| Title | Towards Optimal TTS Corpora |
|---|---|
| Authors | Didier Cadic, Cédric Boidin and Christophe d'Alessandro |
| Abstract | Unit selection text-to-speech systems currently produce very naturalsynthesized phrases by concatenating speech segments from a large database.Recently, increasing demand for designing high quality voices with less datahas created need for further optimization of the textual corpus recorded by thespeaker. This corpus is traditionally the result of a condensation process:sentences are selected from a reference corpus, using an optimization algorithm(generally greedy) guided by the coverage rate of classic units (diphones,triphones, words…). Such an approach is, however, strongly constrained by thefinite content of the reference corpus, providing limited languagepossibilities. To gain flexibility in the optimization process, in this paper,we introduce a new corpus building procedure based on sentence constructionrather than sentence selection. Sentences are generated using Finite StateTransducers, assisted by a human operator and guided by a newfrequency-weighted coverage criterion based on Vocalic Sandwiches. Thissemi-automatic process requires time-consuming human intervention but seems togive access to much denser corpora, with a density increase of 30 to 40% for agiven coverage rate. |
| Language | Speech resource/database |
| Topics | Speech Synthesis, Corpus (creation, annotation, etc.), Speech resource/database |
Full paper  |
Towards Optimal TTS Corpora |
| Bibtex | @InProceedings{CADIC10.608, author = {Didier Cadic, Cédric Boidin and Christophe d'Alessandro}, title = {Towards Optimal TTS Corpora}, booktitle = {Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10)}, year = {2010}, month = {may}, date = {19-21}, address = {Valletta, Malta}, editor = {Nicoletta Calzolari (Conference Chair), Khalid Choukri, Bente Maegaard, Joseph Mariani, Jan Odjik, Stelios Piperidis, Mike Rosner, Daniel Tapias}, publisher = {European Language Resources Association (ELRA)}, isbn = {2-9517408-6-7}, language = {english} } |