Handwritten Digit Recognition System Using Neural Network

I've developed a comprehensive AI project involving a custom neural network in C++ that predicts handwritten digits (0-9) from 28x28 pixel images with an 83% accuracy rate. The project includes a React-based UI for testing using the MNIST dataset and an API built with Python and Flask to facilitate predictions. The API processes grayscale values and outputs the predicted digit. Check out the GitHub repository for more details and setup instructions.