Design and development of cognitive skills enhancement for neurodivergent children using CNN

Addressing the unique challenges faced by neurodivergent individuals, especially young children, this paper introduces CogniQuest-an iOS application designed to enhance basic cognitive skills from an early age. CogniQuest offers three difficulty levels of cognitive tests focusing on memory, object identification, and word processing. The app employs a handwriting recognition CNN (Convolutional Neural Network) model trained on a local dataset from Kaggle using Keras, TensorFlow framework, Open CV and flask for deploying the model using an API. Swift UI is used for processing user input and for rating the handwriting correctness we use the SSIM score. With a minimalistic and neurodivergent-friendly design, CogniQuest ensures accessibility and ease of use. The paper explores the app’s design principles, development process, and the effectiveness of its cognitive enhancement features for the young neurodivergent community.