Could you update the output of the function.

[sunwoo76]

Hi!
I really appreciate your work!
but, It will be good If we can get a detection bounding box also.

Could you add the additional option to get the bounding box of the landmark?

Thanks!

[1adrianb]

Hi, sure I can add a flag for this and return the detected bounding boxes too.
The same can be achieved by calling the face detector

face-alignment/face_alignment/api.py

Line 126 in 2bcfcc6

detected_faces = self.face_detector.detect_from_image(image.copy())

before the face alignment and then passing the bounding boxes to it. But I agree sometimes its more convenient to return the bboxes in one go too.