MPO - Lingea
Official name: Multilingual recognition and search in speech for electronic dictionaries
Official website: www.mpo.cz/dokument50975.html
Abstract:
The proposed project aims at research, development and assessment of technologies for prototyping of speech recognition and search systems with only a few hours of transcribed training data, without the need for phonetic or linguistic expertise. These technologies will be tested in the domain of Lingea electronic dictionaries.
BUT research task:
In this MPO project, BUT builds on its experience with sub-space Gausssian Mixture modelling (GMM) of speech, as it was started at 2009 Johns Hopkins University summer workshop www.clsp.jhu.edu/workshops/ws09/groups/ldchqsrnld/. BUT will extend this work by links to automatically generated pronunciation dictionaries allowing for training a speech recognition system from very limited training data.
| Principal Investigator: | Jan “Honza” Černocký |
|---|---|
| Co-investigators: | Petr Schwarz, Martin Karafiát, Lukáš Burget, František Grézl, Pavel Matějka, Josef Žižka |
Grant agency:
This project is financed by Ministry of Trade and Industry of the Czech Republic, under the programme “TIP”, project number: FR-TI1/034. The project runs from 09/2009 to 08/2013.
Logo:
