Image-to-Text and Speech-based Learning Aid for Children
Ameya Dikshit, Janhavi Bhutki, Ahona Chattopadhyay, Pratham Angre, Nadir Nizar Ali Charniya · 2023
It is well known that kids are more curious and interested to refer to picture-based books and real examples of physical objects for their study rather than traditional books that are merely textual. This is an advantage for teachers and parents and is important for developing their linguistic skills. An application of image-to-text and speech-based learning aid for young children using a deep learning approach is presented in this paper. Using an android application, the image of an object is captured and uploaded on the Firebase, which acts as the primary Cloud Service. The image is then processed and recognized using Convolutional Neural Network (CNN) which yields an accuracy of 93.29% on the test set. The text and speech corresponding to the recognized image are then outputted to the user. The application serves as a learning aid for young children, especially in pandemic-like scenarios.