DL-ASD: A Deep Learning Approach for Autism Spectrum Disorder

Abstract

Identifying a person’s feelings and sentiments is known as emotion recognition and analysis. The emotion analysis approach correctly recognizes normal people’s facial emotions in the first attempt. Children with Autism Spectrum Disorder (ASD) who have trouble talking or expressing themselves can struggle emotionally to understand. To predict ASD and No ASD in children aged 1-10 using dynamic analysis, this work presents a robust deep learning model with multi-label categorization. We proposed a DL-ASD framework for identifying autism spectrum disorder. The proposed model has used the Kaggle dataset as an image dataset. The datasets are trained with an Improved Convolutional Neural Network (I-CNN), and the images are used to classify individuals as having autism spectrum disorder or not having ASD. Feature-based calculations of internal and exterior distances are used to identify the emotion. Optimization procedures such as dropout, batch normalization, and parameter update are used to optimize the Improved Convolutional Neural Network’s (I-CNN) processing of the returning facial landmarks. The proposed method correctly predicts six emotions in addition to four general emotions. According to the experimental results, the classification accuracy of the approach proposed in this study can reach 98%.