Early Childhood Action Monitoring and Analytics System (ECAMS)
Isuru Supasan Naotunna Andarage, Dileepa Fernando, Buddhi Avishka Lokuarachchi, Malithi Gimhani Athuluwage, Pavithra Wijewickrama · 2023
The cultivation of fundamental movement skills (FMS) during the early stages of childhood bears immense importance in shaping an individual’s involvement and achievements in sports throughout their entire life. Therefore, the assessment of motor skills in early childhood (age 3-5 years) significantly influences the advancement of sports within a nation. To this end, early childhood motor skill assessment has been performed manually posing efficiency challenges. The current automation approaches either involve invasiveness or depend on a substantial volume of annotated video training data. In this research, we introduce the Early Childhood Actions Monitoring and Analytics System (ECAMS) toolkit designed to evaluate early childhood motor skills, including activities like sitting up, running, walking, and jumping. This research employs computer vision and machine learning techniques integrated with the TGM2D (Test of Gross Motor Development-2nd Edition) toolkit to accurately detect motor skills from a given video input. This real-time analysis will offer accurate guidance to individuals involved in early childhood development, including educators, trainers, parents, and sports analysts.