Method for speech recognition in patients with neurological disorder or laryngectomy for e.g. vocal rehabilitation, involves acoustically reproducing output signal and/or converting signal into writte
The method involves receiving noise, which is generated in an oral cavity (6) of patients, during articulation of speech of the patients. An input signal is subjected in a digital form in a signal processing device for detecting linguistic patterns. The detected patterns are converted to an output signal by speech synthesis, where the output signal corresponds to natural or simulated articulated speaking voice. The output signal is acoustically reproduced and/or converted into a written text reference number list. Independent claims are also included for the following: (1) a system for speech recognition in patients (2) a method for operating a system for speech recognition in patients.
SCHULDT TOBIAS [DE]; DOMMERICH STEFFEN [DE]; KRAMP BURKHARD [DE]
Type of Offer:
« More Acoustical Patents