SALA II Project Enables GSM Speech Database for telisma

Thanks to the SALA II projects, telisma can now offer a new GSM speech database for South America that recognizes Portuguese and Spanish dialectal variants.

The SALA (SpeechDat across Latin America) II produced a GSM speech database for all American countries, covering all dialectal regions of America and representing the dialectal variants of the English, French, Portuguese and Spanish languages.

A speech database involving the recording of no less than 1000 native speakers has been realized for each linguistic area. Each speaker typically answered a set of questions and read a number of texts via mobile phone or cellular handset, involving the enunciation of isolated digits, phone numbers, credit card numbers, names, last names, regional city names, currencies, major corporations and phonetically rich words and sentences, according to the specifications of the EU-funded SpeechDat project.

As member of the consortium, telisma was granted access to the latest GSM speech database from Venezuela, Chile, Argentina, Peru, Brazil, Colombia and Costa Rica this summer.