Autism Spectrum Disorder Detection of Facial Images Using VGG Model

autism spectrum disorder (ASD) is regarded as a neuro-developmental disorder characterized by difficulties in communication, social interaction, and cognitive func- tioning. It is important to note that early diagnosis of the con- dition enables effective intervention and support; however, classical methods of diagnosis depend on clinical evalua- tions which are often slow and subjective. The purpose of this research is to develop an automated autism behavioral detection system that utilizes MTCNN (Multi-task Cascaded Convolutional Neural Network). The suggested system monitors facial expressions, eye movements, and other body language to determine behaviors typical of autism. This method is very efficient and convenient because it combines modern deep learning methodologies with non-invasive and convenient procedures to allow young children’s ASD detec- tion screening. The goal of the system is to help as many guardians and specialists as possible so they could notify children about the related features of ASD in order to facili- tate swift clinical investigation. The suggested model takes images as input and identifies critical components of the face to ascertain behavioral characteristics of autism. The model classifies and evaluates facial expression photos using deep learning for precise outcome. This system differs from tra- ditional diagnostic procedures in the sense that it lessens re- liance on subjective evaluation and diminishes the resources required for the clinics. Using MTCNN avoids errors in de- tection of facial landmarks caused by traditional MTCNN face features extraction methods.