My master’s thesis at the College of Computer Science and Information Technology at the University of Karbala discussed decision making regarding robot-human interaction based on CNN models.
The study, presented by student Ali Hussein Musa, aimed to develop an Attention CNN model for recognizing emotions, and to present a new method for identifying emotions for a group of individuals.
The study found the possibility of developing an effective Attention CNN model to recognize emotions with high accuracy, and proposed a new method for identifying emotions in groups of individuals based on the distribution of emotions, the size of faces, and their proximity to the robot.
The study recommended the necessity of using a diverse and carefully prepared data set to train future models, and identified aspects of expected limitations and challenges, including video data quality and environments with interactive dynamics.
