Facial feature detection in random data (grayscale) using Python, OpenCV, and numpy.
This is just an experiment based on a random idea I had, it doesn't serve any purpose except for me researching Python.
Run pip install -r requirements.txt
to install the dependencies.
Run python3 app.py
and wait until a window with an image shows.
An option to save generated images (only ones containing "faces") instead of showing them is available: python app.py -s
The code runs single threaded, don't expect great speeds. It took around 5 minutes to find the first face, although your mileage may vary.
I'm getting speeds that hover around 60-70 images generated per second. Python-OpenCV runs on the CPU so the speeds will depend only on how powerful the CPU is.