A Review On: Autism Spectrum Disorder Detection by Machine Learning Using Small Video

Abstract

Autism spectrum disorder (ASD) is a mental ailment that can be diagnosed by the study of social media data and biopsy. Those with autism spectrum disorder (ASD), a neurodevelopment condition, may experience permanent changes to their facial appearance over time. The faces of children with ASD are easily identifiable from those of normally developing (TD) children. After the toddler years, specialists will typically look at a child's behaviour patterns to make a diagnosis of autism spectrum disorders (ASD). Quicker intervention and better long-term outcomes are possible after an early diagnosis of autism spectrum disorder. Machine learning uses data science to facilitate early autism diagnoses. This literature review aims to bridge a gap in understanding by bringing together the results of recent studies and technologies that use machine learning based approaches for ASD screening in infants and children younger than 18 months. Individuals on the autism spectrum have restricted interests and behaviors and struggle to communicate and interact socially with others. The prevalence of ASD has increased in recent years. The potential for innovative methods like machine learning to be integrated into established therapeutic practices is quite encouraging. How computers can be taught to recognize patterns in data is the focus of machine learning research. Artificially intelligent technologies can identify signs and symptoms, organize data, make diagnoses, and forecast outcomes. Machine learning is an area of artificial intelligence that focuses on teaching computers new behaviors by observing how they interact with the world. These days, there are a wide variety of machine learning methods available. In this piece, we look at the existing literature on the frequency of ASD in the general community. The main interest of this paper is the academic literature were sought by searching numerous databases.