Efficient Classification Of Autism in Children based on Resnet-50 and Xception module

Children with autism spectrum disorder struggle to process emotions appropriately, which makes it difficult for them to understand facial expressions. It is crucial for autistic children to recognize emotions because emotion is a crucial component of good socialization. In the proposed method, we use the two different CNN architecture which includes Resnet-50 and Xception modules, and designed an app to classify autistic and non-autistic children. The results demonstrated that the accuracy of the Resnet-50 method is better when compared to the existing conventional methods.