EduRobot: Intelligent Humanoid Robot with Natural Interaction for Education and Entertainment

Abstract

This paper presents a research on intelligent humanoid robots for education focused on natural interaction for teaching basic mathematics and entertainment for students. Previously, humanoid robot focused on the architecture of body and ethical issues, but later, it is predicted that humanoid robots should consider teaching behaviour, student's interest and speech understanding. They are expected to conclude the answer through conversation under noisy environment. We propose speech recognition systems based on Google translator and basic Natural Language Processing (NLP) techniques such as stemming and tokenization. We do teach modules for elementary school in order that the robot can interact with students naturally based on psychological aspects. We also discuss about our recent humanoid robot for education in elementary school that has ability to detect user's face and accepts commands in sentences from the user to do any actions such as giving tutorial and answering questions. The average accuracy for correct recognition is 73.3% and correct answer is 63.3%. Our speech recognition system is more promising compared with NAO robot.