Content area
Full Text
ABSTRACT: The Smart Learning Tool for Kids with Real-time Image Classification is an Al-powered educational tool designed to assist young learners in recognizing and identifying objects through real-time image classification. The system captures images using a webcam, processes them through a Convolutional Neural Network (CNN) model and outputs the corresponding class label. It provides immediate audio feedback by pronouncing the class name in four languages: English, Myanmar, Thai and Chinese. The system aims to enhance kids' learning experience by engaging multiple senses-visual and auditory-that makes learning interactive and multilingual. The CNN model is trained with custom training data, enabling accurate classification of 12 object classes. This system serves as a smart and user-friendly tool for early childhood education.
KEYWORDS: Real-time Image Classification, Convolutional Neural Network (CNN), Multilingual Audio Feedback, Early Childhood Education, Interactive Learning, Al in Education.
1 Introduction
Technology is advancing rapidly, transforming many aspects of our lives, including education. The integration of modern technology into learning environments offers significant advantages. One major benefit is the ability to create more engaging and interactive learning experiences. Advanced tools and systems can make learning more enjoyable and effective by providing real-time feedback and personalized support.
Technology also supports cognitive growth by aiding in skills like recognizing objects and learning languages. Tools that support multiple languages and interactive materials help make education more accessible and inclusive for everyone regardless of their background.
According to [1], computer vision has become increasingly intriguing in recent years, particularly with the rise of self-driving cars. Generic object detection focuses on identifying and classifying objects within a single image, marking them with rectangular bounding boxes and indicating their confidence levels. Today, the advancement of object detection has been significantly enhanced by deep convolutional neural networks (CNNs), which have recently become the leading approach in object recognition research due to their exceptional performance on various challenging datasets, like ImageNet, gathered from the web.
In exploring the role of Al in education, [2] provides a comprehensive overview of how Al can enhance personalized learning and address ethical concerns like data privacy. Similarly, in [6], the authors focus on designing Al-driven educational systems that cater to the evolving needs of learners. Both books offer valuable insights into the integration of Al in educational environments,...