Detecting faces with deep neural nets and Caffe

One advantage of using OpenCV's implementation of visual neural networks is that they are available on different platforms. For the sake of clarity and brevity, we are using Python on an environment that has Python installed. However, the same results could be used with OpenCV's C++ implementation on an ARM-CortexM3 or OpenCV's Java implementation on an Android system. In this recipe, we are going to use a face detection neural network that OpenCV implemented based on the Caffe machine learning framework. The output of this recipe will be a window on the PC that has the image with bounding boxes around the face.