Home

Purpose of this Wiki

This wiki serves to provide information on the two distinct aspects of this project: the model, and the flask web application, and to provide sample test screenshots of each digit being recognised by the mode. Please visit the relevant page for each to read up on how they work individually and how they work together:

• Model
• Flask Web Application
• Testing the Web Application

The Project

This project concerns the well-known MNIST dataset and the Python packages Keras, Flask, and Jupyter. This project contains a jupyter notebook containing code to train the model, and a flask web application that allows users to recognise digits using the trained model.

References

• Image Classification in 10 Minutes with MNIST Dataset
• Deep Learning Notebook
• Drawing on HTML5 Canvas using a mouse
• Python: Resizing and Fitting an Image to an Exact Size
• Python PIL | Image.point() method
• Bug fix: expected conv2d_1_input to have shape (28, 28, 1) but got array with shape (1, 28, 28)
• Dense layers