NumPy-Complete is a neural network to recognise handwritten digits, trained using only NumPy(No TensorFlow, Pytorch) on the MNIST database.
- Make sure the following packages are installed:
pip install numpy pandas matplotlib opencv-python- Clone the repository
git clone https://github.com/fringewidth/numpy-complete.git-
Modify
input-image.pngto a custom handwritten digit. The image must be 28px$\times$ 28px -
Run
run-model.py
- Recognises a handwritten digit from a 28
$\times$ 28 pixel grid using a feed-forward neural network with two hidden layers, with an accuracy of 90.41%. - Makes use of only NumPy for numerical processing. All functions ar custom implemented and modifiable.
- Ability to run custom input.
- Training notebook available for customizing the training process. See train.ipynb for details on architecture and implementaion.