Speech resources

  • Brøndsted, Tom (Projektdeltager)



Speech and Multimedia Communication A number of (Danish) speech resources including both databases and software modules are maintained. The resources encompass speech recognition software (the large vocabulary speech recogniser Slang, ported versions of HTK 3.0, java-interfaces to HTK), acoustic models to be used with the speech recognition modules (models for Danish and other languages trained on the SpeechDat-II database), Java-interfaces to MSS API compliant speech synthesizers, and a NLP suite for speech understanding. For details, see http://cpk.auc.dk/~tb/speech. (Tom Brøndsted)
Effektiv start/slut dato31/12/200131/12/2001