face detection - yup
face recognition - sorta. Currently only trained to recognize Beyonce.
face greening - incomplete. Lazy implementation only paints entire square green. Needs something to do background subtraction to isolate face.
eye detection and morse code from blinks - eyes are detected and found but morse code inconsistent
tested on Ubuntu 16.04
checkout: git clone [email protected]:MisterYu:faces.git
go into checkout directory: cd faces
create virtual environment: python3 -m venv env_face
activate environment: . env_face/bin/activate
install requisite modules: pip install -r requirements.txt
install faces: python3 setup.py install
activate environment: . env_face/bin/activate
python3 faces.py
to bring up the gui in py_faces
directory. Paths hard coded relative (sorry).
everything is off by default
click "start" to have camera start collecting data
click "stop" to have camera stop collecting data
click "Face Detect" to start face detection. "Eyes?", "Hulk out", and "Beyonce Recognition" won't work without "Face Detect"
click "Eyes?" to start eye detection and Morse code deciphering from blinks
click "Hulk out" to turn box with face in green; would be better if only face greened
click "Beyonce Recognition" to kick up Queen B recognition. Adjust "confidence" levels
test code - NOPE. need more time to write all the test code. TDD fail.
C++ inclusion - NOPE. all time burned on feature implementation. Ideally use C++11 to thread face recognition on more than Queen B.
In order for modifications to show up, the modified ui file requires the following processing: pyuic5 faces_gui.ui -o faces_gui.py