Speech databases covering 18 languages will be created. The majority of the European languages, as well as the Chinese and Japanese languages, are part of the SPEECON project.
If you are interested in learning more about the Speecon project, you may consult the Speecon Web site at the following address: www.spechdat.org/speecon/index.html
ELDA is responsible for the French recordings and is in charge of supervising the recordings for the Italian and Swedish languages. For the three languages (French, Italian & Swedish), 600 speakers, distributed in 6 dialectal regions (in each country), will be recruited. The recordings will then take place in 5 different environments (entertainment, office, public place, car, children room).
ELDA will provide the linguistic data to be used for the development of voice-controlled products.
Start Date: February 2000 (the speech data collection only started from June 2001)
Duration: 3 1/2 years
Khalid Choukri, ELDA CEO at
- Daimler Chrysler
- Lernout & Hauspie
- Philips Speech Processing
- Temic Telefunken
- Natural Speech Communication
- Nordisk Spräkteknologi