Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/75241B19-EC58-4BE9-934A-B92ED3E9A10B | ||
| Year | 2025 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Efstathia-Maria Chatziathanasiou, "Development of robot tutor for kids", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2025 https://doi.org/10.26233/heallink.tuc.104956 | ||
| Appears in Collections |
In our days, artificial intelligence has significantly contributed to the advancement of technology. It has introduced new scientific fields, one of the most important being machine learning, which focuses on the study and development of algorithms capable of learning from data and making predictions based on it. Machine learning has played a major role in improving people’s daily lives, directly influencing their social life and, more specifically, their education. Motivated by the above, we implemented this diploma thesis, which focuses on developing a simple robot tutor for children, using the humanoid robot NAO. In the proposed setup, the robot interacts through two simple games with a child writing on a board. In the first game, the robot helps the child learn the correct spelling of words that describe everyday objects. In the second game, the robot helps the child correctly evaluate simple math expressions, which include addition, subtraction, multiplication and division. During the execution of this work, we tackled the problem of visual object recognition and the problem of visual recognition of handwritten text. With the help of the OpenCV library, we successfully retrieved real-time images from the robot camera, and by using the YOLOv3 model, we completed the object recognition process. Next, we faced the challenge of recognizing handwritten text, where we applied appropriate image processing algorithms from OpenCV to improve the image quality. Finally, using properly trained neural networks, we successfully resolved the text recognition task. Our robot tutor was evaluated by a small group of pupils just finishing the elementary school. This project was positively evaluated, as the NAO responded as expected, making most of the time the correct decisions based on each child’s input. The results demonstrate the effective interaction between NAO and the children, indicating the significant impact of integrating robots into human education.
Copyright © DIAS 2013
