A simple implementation of facial detection using the pretrained Caffe
model.
- the
prototxt
file contains the model architecture i.e. the layers - the
ResNet caffemodel
file contains the weights of the actual layers of the SSD (Single Shot Detection) model.
on images:
python detect_faces.py --image your-image-file.jpg --prototxt deploy.prototxt.txt \
--model res10_300x300_ssd_iter_140000.caffemodel
on webcam:
python detect_faces_video.py --prototxt deploy.prototxt.txt --model res10_300x300_ssd_iter_140000.caffemodel