digit-recognition-using-cnn

0 0
  • 0 Collaborators

Implementation of CNNs for handwritten digit recognition ...learn more

Artificial Intelligence

Groups
Student Developers for AI

Overview / Usage

Digit recognition using convolutional neural network

Overview

A web application that recognizes handwritten digits. This application has been made using

TensorFlow
Keras
Flask
This model was trained using the MNIST Handwritten digits dataset.

To update tensorflow and keras to their latest version use this

$ sudo pip install tensorflow -upgrade
$ sudo pip install keras -upgrade
Acquiring the repository

$ git clone https://github.com/prajjwal1/convolutional-neural-netwok.git
$ cd image_recognition_using_cnn_web_app
Dependencies

$ sudo pip install -r dependencies.txt
To run the server use this:

$ python app.py
Note the address and open it on your browser. Press ctrl-c to terminate

Traning the model

To train the model for yourself:

$ python train.py
It's recommended that you use Tensorflow with gpu support (CUDA).

Comments (0)