https://github.com/edanea/cnn-digits
Handwritten digit categorization model with React application for interacting with model.
https://github.com/edanea/cnn-digits
convolutional-neural-networks digit-recognition jupyter-notebook keras machine-learning react tensorflow tensorflowjs
Last synced: 2 months ago
JSON representation
Handwritten digit categorization model with React application for interacting with model.
- Host: GitHub
- URL: https://github.com/edanea/cnn-digits
- Owner: EdanEA
- Created: 2025-02-27T00:20:44.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-03-12T08:23:23.000Z (over 1 year ago)
- Last Synced: 2025-07-20T12:24:46.873Z (12 months ago)
- Topics: convolutional-neural-networks, digit-recognition, jupyter-notebook, keras, machine-learning, react, tensorflow, tensorflowjs
- Language: Jupyter Notebook
- Homepage: https://edanea.github.io/cnn-digits/
- Size: 49.9 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Handwritten Digits CNN
A simple model trained on handwritten and font digits, 0-9. The dataset is a modified public dataset from Kaggle: https://www.kaggle.com/datasets/pintowar/numerical-images.
A React application was made for using the models trained created with the notebook. The source for the React portion can be found in the [react](https://github.com/EdanEA/cnn-digits/tree/react) branch. A deployed version of the React application is also at https://edanea.github.io/cnn-digits.
## Usage
1. `git clone https://github.com/EdanEA/cnn-digits.git`
2. `cd cnn-digits`
3. `jupyter notebook "Digit CNN.ipynb"`
4. Run all cells in notebook for dataset to be untarred, processed, and then have the CNN compiled and trained.
## Requirements
* Python <= v3.12
* pip >= v24.2
* Jupyter Notebook >= v7.0
All other required packages will be installed when running the notebook.