https://github.com/salvacorts/keras-mnist-paint
:bar_chart: An interactive GUI to draw numbers and recognize them using a CNN written with Keras and trained with MNIST
https://github.com/salvacorts/keras-mnist-paint
draw-numbers gui keras machine-learning mnist neural-network paint python tkinter
Last synced: 2 months ago
JSON representation
:bar_chart: An interactive GUI to draw numbers and recognize them using a CNN written with Keras and trained with MNIST
- Host: GitHub
- URL: https://github.com/salvacorts/keras-mnist-paint
- Owner: salvacorts
- Created: 2017-12-26T14:28:04.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2017-12-26T21:02:01.000Z (almost 8 years ago)
- Last Synced: 2025-07-11T18:49:13.798Z (3 months ago)
- Topics: draw-numbers, gui, keras, machine-learning, mnist, neural-network, paint, python, tkinter
- Language: Python
- Homepage:
- Size: 4.48 MB
- Stars: 10
- Watchers: 1
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Interactive MNIST Paint GUI
This is a PoC of using Keras to build, train, save and put in into production a model.
I'm just beginning with Machine Learning so probably this is not the best model.
The interface offers a canvas square to draw numbers in and gives the predicted number within the scores for the rest of numbers.
## Requirements installation
- Pyhton 3
- Tkinter: `sudo apt install python3-tk`
- Tensorflows: https://www.tensorflow.org/install/install_linux
- Keras, Matploitlib, numpy: `pip3 install -r requirements.txt`
## Running
Clone this repository and move there:
`git clone https://github.com/salvacorts/Keras-MNIST-Paint.git && cd Keras-MNIST-Paint`
Run the program:
`chmod +x main.py && ./main.py`
## Hacking
Although this project comes with a pre-trained model, modifying and running [modules/train.py](https://github.com/salvacorts/Keras-MNIST-Paint/blob/master/modules/train.py) you can train your own model and tweak this Neural Network.