What is face detection NN output looks like?

Course Q&A Deep Learning Specialization Structuring Machine Learning Projects
1

Hi Mentors and friends,

I just done the section of " End-to-end Deep Learning". In this section, Prof mentioned about face detection. So, you have a picture, the NN will find the section (kinda make a square around the face) of the face, and zoom in.

If this is truly done by NN, may I ask, what’s the output of this NN look like? Because of what I learned so far, the output is just a yes or no label (e,g, yes or no it’s a cat, yes or no there is a car in the pic). but in this case face detection, the output should be a location (like X, Y ?)

I understand this is a little bit out of the scope, so I just need a very basic concept here, a tutorial link is good as well.

2

NNs can do more than just classifications. I suggest you “hold that thought” and take Course 4 in the series. That includes Face Recognition and YOLO, which can draw bounding boxes around objects in a scene and identify the objects. Both of those networks are “convolutional networks”, which is the overall subject of Course 4.

3

oh, nice. I dont know Course 4 will take about that. haha. thank you!