COMBINED APPROACH OF SPEECH INTELLIGIBILITY FOR CALMING CHILDREN WITH SPEECH DISORDERS

Evaluating the degree of dysarthria's severity can help pathologists plan therapy, help automated dysarthria speech recognition systems, and give insight into how well the patient is improving. This article presents a comparative research on the use of several deep learning algorithms and acoustic characteristics for the categorization of dysarthria severity levels. First, we assess the fundamental architecture options, including the convolutional neural network, DNN, GRU, and LSTM, utilizing fundamental characteristics Subsequently, DNN models are used to assess aspects related to speech disorders