Method to adapt dynamically the acoustic models of recognition of the speech to the user. (Machine-translation by Google Translate, not legally binding)

摘要

Method for dynamically adapting acoustic models of speech recognition to the user, which allows improving the efficiency of speech recognition systems used by tele-operators (or any service that requires speech recognition through telecommunications networks) and that it is based on the sending of data of the user's speech profile, obtained in a single previous training and valid for all subsequent sessions in any of speech recognition systems compatible with this method, adapting the acoustic models to the speaker to improve the rate of speech. Success of such systems and avoid the user the need to train multiple platforms. (Machine-translation by Google Translate, not legally binding)