Just run capture-detect.py
to extract all the faces detected with the camera with landmark extraction and extra padding.
- This is not the whole system, only the face detection part. The rest is implemented in MAX.
- An earlier version of this project was presented in ESCALATRONICA 2015, Buenos Aires, as Imaginatu.
- Depends on:
- OpenCV 2.4
- Dlib 18.8
- Python packages: numpy and skimage
- Create a directory
entities/
to store the data, or change this in thecapture-detect.py
script - Create a
data/
directory and put theshape_predictor_68_face_landmarks.dat
data file from dlib there