Automatic Prosodic Analysis to Identify Mild Dementia

摘要: This paper describes an exploratory technique to identify mild dementia by assessing the degree of speech deficits. A total twenty participants were used for this experiment, ten patients with a diagnosis and like healthy control. The audio session each subject was recorded following methodology developed present study. Prosodic features in elderly controls measured using automatic prosodic analysis on reading task. novel method carried out gather twelve over samples. best classification rate achieved 85% accuracy four features. results attained show that proposed computational offers viable alternative identification adults.