Adapting voice recognition for the vocally disabled
Verbal communication is something that most people take for granted. Deterioration of sight and hearing may be more common ailments, but the loss of one's voice can be just as devastating. Many professions depend heavily on the use of one's voice, such as teachers, singers, etc. Human vocal cords can be damaged by diseases, diet, overuse and environmental factors. What is needed is a methodology for assessing individual voice risk. A project funded in part by the Telematics 2C Programme satisfied this need by establishing a predictive scale for vocal risk. Subjects complete a short questionnaire consisting of just twenty-five yes-no questions. Different risk factors are assessed, including vocal load, lifestyle, general and vocal health. The final score on the questionnaire is used to rank individuals on the risk scale. Scores are group into categories such as "heavy load", "normal load" and so on. The new methodology was tested on a sample population of 26 individuals. A final report analyses the results of the trial and provides concrete recommendations for improved vocal care. These guidelines will help teachers and other high-risk professionals reduce the chances of voice impairment. Another important group that can benefit from this work is the motor-impaired or disabled individuals. The predictive voice risk scale and associated methodology will contribute to the development of superior voice recognition software, helping them live happier, more productive lives.
