This is a small fork of activatedgeek/LeNet-5 by Ciro Santilli adding better integration and automation for:
- extracting MNIST images as PNG
- ONNX CLI inference taking any image files as input
- a Python
tkinterGUI that lets you draw and see inference live - running on GPU
Install on Ubuntu 24.10:
sudo apt install protobuf-compiler
cd lenet
virtualenv -p python3 .venv
. .venv/bin/activate
pip install -r requirements-python-3-12.txtDownload and extract MNIST train, test accuracy, and generate the ONNX Extract MNIST images as PNG:Infer some individual images using the ONNX:Draw on a GUI and see live inference using the ONNX:TODO: the following are missing for this to work:
lenet.onnx:./train.py./extract_pngs.py./infer.py data/MNIST/png/test/0/*.png./draw.py- start a background task. This we know how to do: stackoverflow.com/questions/1198262/tkinter-locks-python-when-an-icon-is-loaded-and-tk-mainloop-is-in-a-thread/79502287#79502287
- get bytes from the canvas: all methods are ugly: stackoverflow.com/questions/9886274/how-can-i-convert-canvas-content-to-an-image
New to topics? Read the docs here!