Intel® Distribution of OpenVINO™ Toolkit
Community support and discussions about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all things computer vision-related on Intel® platforms.

What is the output of facial landmarks model

R__A
Beginner
149 Views

Hi

I am working on facial landmarks model. It is perfectly working and drawing all the 35 points over the face. As per the below page:

http://docs.openvinotoolkit.org/latest/_facial_landmarks_35_adas_0002_description_facial_landmarks_3...

it says,
The net outputs a blob with the shape: [1, 70], containing row-vector of 70 floating point values for 35 landmarks' normed coordinates in the form (x0, y0, x1, y1, ..., x34, y34).

Does this mean the model will output the coordinates of all the landmark points over face.?

Lets say, there is a side face for which one of the eye is not visible, so in this case model will still show some coordinates value or it will show null value.

My main objective with this model is to check if any of the face landmark shows null value i.e left eye value is null which means the face is side face. Is this possible with this model.?

Thanks

0 Kudos
2 Replies
Sergei_N_Intel
Employee
149 Views

> Lets say, there is a side face for which one of the eye is not visible, so in this case model will still show some coordinates value or it will show null value.

There won't be any indication of missing points. The model will assume that all of the points are visible and will try to "make its best" in case some points are actually missing. It's up to the application to decide what to do in such cases. 

R__A
Beginner
149 Views

Thanks for your late reply I got it working on my own.

Reply