A Deep Convolutional Neural Network based Detection System for Autism Spectrum Disorder in Facial images

Autism spectrum disorder (ASD) a development disability which causes several challenges in social, communication and behaviour abilities. ASD people are nowhere much different when compared to ordinary people rather the way they interact will be different, few ppl of ASD needed help for all basic needs and others don’t. On effectively identifying ASD at the earlier stages helps to provide therapy to improve their skills. Being a disability in neurological development, many researchers are trying to predict the ASD in advance with Image processing techniques based on MRI Images. This research work has attempted to develop a prediction system based on Convolution Neural Network [CNN] based on their photos. Database for the required model is taken from Kaggle and split into 80:20 for training and testing the model. Our model managed to give an accuracy rate of 91% and a overall loss of 0.53.