AI‑Driven Assistive Application for Early Childhood Education
Year: 2026
Type: report
Abstract: This work presents the design, development, and evaluation of Kidzinya, an AI-driven assistive application for early childhood education. The application provides an interactive learning environment that combines visual, audio, speech, writing, counting, and object-based learning activities while allowing guardians to manage child profiles and monitor learning progress. The system integrates multiple artificial intelligence and computer vision technologies, including YOLO-based object detection, Whisper-based child speech recognition, handwritten letter and digit recognition using convolutional neural networks, finger-counting recognition, Tesseract OCR, and Piper text-to-speech. The project evaluates these models using quantitative performance metrics as well as their practical suitability for deployment in a child-facing educational application. The report covers the theoretical background, literature review, system requirements, architecture and design, AI model development, dataset preparation, testing procedures, experimental results, system implementation, and future improvements. This work was developed as a Junior Project at the Faculty of Informatics and Communication Engineering, Arab International University (AIU).
Author: Abdeen, Aya
Institution: Arab International University
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7203761990