Emotion Recognition Technology for People With Autism: A Systematic Literature Review

With the rapid growth and availability of smart technology, the idea of households using smart technology (smartphones, computers) is becoming popular. This availability of smart technology inside homes can potentially help with critical problems such as understanding the emotions of people with autism. Despite their high potential in solving complex problems, such technological solutions are largely unexplored. We take a first step in this direction by conducting a systematic literature review using three major databases: Google Scholar, ACM Digital Library, and IEEE Xplore. Our findings indicate that most of the existing approaches only utilize facial expressions for model development and collect data from children with autism, indicating a need for utilizing technology to capture more than facial expressions, capture contextual and other available information, and collect data from a diverse group of individuals with autism as it will improve the effectiveness of the machine learning models.