Watch a neural network classify your fruit drawings. GUI was made using Electron.js paired up with a Python backend running a flask server.
Watch the video here : https://youtu.be/O8Gvkq8y-qs
Google colab notebook : https://colab.research.google.com/drive/1pG7gbXyAq-8UL_Pj18FV6O-jU0MmYZze
- Python
- Electron.js
- node.js
- Tensorflow
- Keras
- numpy
- scikit-learn
- PIL
- scipy
- Flask
Install electron (You'll need node.js)
$ npm install electron -gDownload and move into the project directory
$ git clone https://github.com/SouravJohar/doodle-classifier.git
$ cd doodle-classifier
Download the dataset inside a 'data/' directory. Dataset : https://github.com/googlecreativelab/quickdraw-dataset (get the numpy bitmaps)
$ python train.pyAfter training, link your saved model in server.py and then run the server
$ python server.pyAfter the server is up and running,
$ electron .