<![CDATA[Children with autism often have difficulties in social interaction and communication. Socially Assistive Robots (SAR) can be utilized to help them develop social and communication skills. One of the important aspects in interaction with robots is the ability of the robot to detect the motion of autistic children. This research aims to implement the Convolutional Neural Network (CNN) method on the SAR robot to detect the motion of autistic children. The CNN method was chosen because of its ability to recognize image patterns and detect objects with high accuracy. The image dataset used in this study consists of images taken using a cellphone camera manually with various poses from various points of view. The results show that the CNN method is able to recognize the movements of autistic children with high precision, thus supporting more effective interaction between the robot and the child. Challenges such as large dataset requirements and significant training time were overcome through optimization of the training process. This optimization led to improved model efficiency, reducing computation time while maintaining high accuracy.]]>
Copyrights © 2025
