Hand Writing Recognition Using Convolutional Neural Networks

Introduction

This CNN-based model for recognition of hand written digits attains a validation accuracy of 99.2% after training for 12 epochs. Its trained on the MNIST dataset on Kaggle.

##Usage The model architecture and weights are saved in the files model_architecture.json and model_weights.h5. Note that these weights are compatible only with the Tensorflow backed.

To train the model run train.py. The file test.py generates a file predictions.csv which contains the predicted labels to the images in the test set. This file can be used for submission at Kaggle. display_random.py displays 25 random images from the test set along with their predicted labels.

Requirements

• Python 2.7
• Tensorflow
• Keras
• h5py
• numpy
• matplotlib
• pandas

Dataset

• The model is trained on the MNIST dataset downloaded from Kaggle.

• The file train.csv contains pixel intensity values as flattened vectors for 42000 images and their corresponding labels. Similarly, test.csv has pixel intensity values for 28000 unlabelled images.

The Model