Image Explorer

Abstract

Today, Modern Technology is taking over the field of education, serving as an essential teaching tool for facilitating student knowledge acquisition. It is found that students, especially young learners, respond more effectively to picture-based books and actual physical objects for their study instead of traditional books that are primarily textual. This is very beneficial for teachers and parents in developing student’s linguistic skills. In this paper, an application of image-to-text and speech-based learning aid for young children using a deep learning approach is presented.

The system will enable the children to point the camera at a specific image to convert it to text, and corresponding speech. It also returns some information and sample image about that object. The app is developed using Kotlin language. Google Firebase is used to connect the app and the server. An Efficient B1 algorithm is used to identify the name of the target image. It is trained with more than 1000 images of each object to predict the results with 93.1% accuracy.

The application aims to revolutionize learning aids for young children, especially in pandemic-like scenarios and make them more effective and intuitive.