dnn facedetection

1. SSD

opencv's uint8 model is only 2.6 mb !
and it is as straightforward as it looks:

	dnn::Net net = dnn::readNetFromTensorflow("opencv_face_detector_uint8.pb","opencv_face_detector.pbtxt");
	Mat res = net.forward("detection_out");

	Mat faces(res.size[2],res.size[3], CV_32F, res.ptr());
	for (int i=0; i(i);
	    float batch = data[0];
	    float confidence = data[2];
	    if (confidence > confThreshold)
	    {
	        int left = (int)(data[3] * f.cols);
	        int top = (int)(data[4] * f.rows);
	        int right = (int)(data[5] * f.cols);
	        int bottom = (int)(data[6] * f.rows);
	        int classId = (int)(data[1]) - 1;  // Skip 0th background class id.
	        ...

2. MTCNN (using opencv's dnn module)

a 3 step process, using 3 seperate (caffe) networks:

step1 step2 step3

1. grab detection candidates from an image (scale) pyramid
2. refine boxes regression
3. derive 5 principal face landmarks (eyes, nose, mouth corners)

3. PCN

original caffe code and model from here