MNIST (Modified National Institute of Standards and Technology) is a well-known dataset used in Computer Vision that was built by Yann Le Cun et. al. It is composed of images that are handwritten digits (0-9), split into a training set of 50,000 images and a test set of 10,000 where each image is of 28 x 28 pixels in width and height.
This dataset is often used for practicing any algorithm made for image classification as the dataset is fairly easy to conquer. In this project I built a convolutional neural network using Keras which achieves 99.8% accuracy on the dataset and used streamlit and heroku to deploy the model.
Web interface to test the app:
https://mnist-cnn-lv.herokuapp.com
- Streamlit : https://www.streamlit.io/
- Deploying the app on Heroku: https://towardsdatascience.com/deploy-streamlit-on-heroku-9c87798d2088
- Show your ML Project to the Internet in Minute: https://towardsdatascience.com/show-your-ml-project-to-the-internet-in-minutes-2a7bc3167bd0
- Super easy to use template: https://github.com/patryk-oleniuk/streamlit-heroku-template/generate (just click the button on the top-right and name your app)