This is fun art.
OpenCV is used to reflect back the eyes of the watcher - Scooby Doo style.
Credits:
- This is inspired by, and re-uses code from Real-time facial landmark detection with OpenCV, Python, and dlib by Adrian Rosebrock
- Uses Dlib's face detection model, trained on data from Imperial College London - C. Sagonas, E. Antonakos, G, Tzimiropoulos, S. Zafeiriou, M. Pantic. 300 faces In-the-wild challenge: Database and results. Image and Vision Computing (IMAVIS), Special Issue on Facial Landmark Localisation "In-The-Wild". 2016. Note: this data set not for comercial use
If you know Python, then you may have your own way of getting setup. I suggest using Miniconda Python because it will download the prebuilt libraries required.
A lot of stuff gets download, so it might take a while, but it should be simple.
Download Miniconda3 4.5.4. This installs Python 3.6.5 which is known to work (a dependency is broken in Miniconda 4.7.12 - latest at Nov 2019).
Run the installer and accept all the defaults.
Find and run "Anaconda Prompt" from the Windows menu, and skip down to "Install required packages".
Download the relevant Miniconda3 4.5.4 (not Miniconda2) from https://repo.continuum.io/miniconda/.
Install by muddling through https://conda.io/projects/conda/en/latest/user-guide/install/index.html.
At the "Anaconda Prompt" enter:
cd <path-to>\scooby_doo_eyes
conda env create --prefix ./env --file environment.yml
You may be asked if you are sure - enter 'Y' and press enter. This bit takes a while.
At the "Anaconda Prompt" enter:
cd <path-to>\scooby_doo_eyes
env\python fetch-model.py
At the "Anaconda Prompt" enter:
env\python scooby_doo_eyes.py
Quit by pressing ESC key. Display the intermediate stages by pressing 0, 1, 2 etc..