Handwritten-digit-recognition-MNIST

Handwritten Digit Recognition using Convolutional Neural Networks in Python with Keras

MNIST dataset:

MNIST is a collection of handwritten digits from 0-9. Image of size 28 X 28

Code Requirements

python 3.x with following modules installed

1. numpy
2. seaborn
3. tensorflow
4. keras
5. opencv2

Description

This is a 5 layers Sequential Convolutional Neural Network for digits recognition trained on MNIST dataset. I choosed to build it with keras API (Tensorflow backend) which is very intuitive.

It achieved 98.51% of accuracy with this CNN trained on a GPU, which took me about a minute. If you dont have a GPU powered machine it might take a little longer, you can try reducing the epochs (steps) to reduce computation.

Execution

To run the code type,

python digit_recogniser.py

Tutorial

Note: This page is not complete. Sorry for delay.

Need help for this to be completed

For step-by-step tutorial please refer to wiki. It will take you through all the steps right from loading the data to recognising digits through live cam.

Update

For running on GPU enabled devices:

Please uncomment the following line from digit_recogniser.py (line no. 70) file:

tfback._get_available_gpus = _get_available_gpus

Note: If you are using the tensorflow 2.1, then you may get an error "AttributeError: module'tensorflow_core._api.v2.config' has no attribute 'experimental_list_devices'"

As the experimental_list_devices is deprecated in tf 2.1. A simple snippet is injected into the code to make the code work. And the code is taken from here : keras-team/keras#13684 (comment)