1.
Noothi Sravan Kumar
– Computer Science And Engineering, St. Peter Engineering College, Hyderabad, Telangana, India.
2.
P. Vamshi Krishna And Chirra Anil
– Computer Science And Engineering, St. Peter Engineering College, Hyderabad, Telangana, India.
Abstract
Social engagement, the manifestation of interpersonal relationships during interaction, is a measure of people’s interest in that interaction. One of the most important challenges in human-robot interaction (HRI) is measuring social engagement, which is necessary for understanding interaction patterns and enabling robots to adjust their behaviour appropriately. The main objective of this study was to advance the theoretical literature and related concepts of social engagement. Developing a trustworthy neural network model for the automated assessment of social engagement was the second objective. Using the PInSoRo dataset, a multilayer perceptron (MLP) classifier was developed and trained to detect social engagement states. Once the model parameters were carefully adjusted, the evaluation demonstrated excellent performance with an accuracy rate of 94.85%.
Keywords Adaptive robotics, Interaction between humans and robots, Machine learning, Measurement of social engagement, Neural systems, Social robots, User involvement.
