Abstract:Children with Autism Spectrum Disorder (ASD) oftentimes have difficulty in both interpersonal and intrapersonal interactions. Most children with ASD lack emotional intelligence and cannot properly comprehend social cues and responses of the people around them. These social impairments are causing them to have difficulty in maintaining relationships thus, to combat these issues, the researchers presented an application with Facial Emotion Recognition (FER) technology using the deep-learning approach namely Convolutional Neural Network (CNN). Using the FER allows the children to perceive the emotions of their interaction partner at face value. The model used in developing the application sits at an accuracy level of 98%. In addition to the FER technology, the research implements a kind of mood meter that enables the children to plot their mood and a dashboard that tracks the accumulated plots. The tracker lets the children convey their emotions as well as help their interaction partner to understand them better. Furthermore, the application includes a recommendation system that suggests various relaxation and meditation techniques that the child can use to manage outbursts or spontaneous emotional activity. The Convey mobile application aimed to empower children that has ASD and help them lead a normal life.

Convey: Developing a Facial Emotion Recognition System for Enhancing Interpersonal Interactions of Children with Autism Spectrum Disorders