Emotion Recognition in Children with Autism Spectrum Disorder Using Convolutional Neural Networks

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder, and it is defined as the persistent difficulty in maturing the socialization process. By applying different therapies, health professionals have made advances that promise to improve patients’ conditions. Taking advantage of technologies like Artificial Intelligence (AI) techniques, improvements in therapies have been obtained. This article proposes creating a smart mirror to recognize five basic emotions: Angry, Fear, Sad, Happy, and neutral. The above is based on Convolutional Neural Networks (CNN), which can support therapies performed by health professionals to children with ASD.