Closed Jusya closed 7 years ago
Hi,
This should be fairly easy since the call to detect_face
:
bounding_boxes, _ = align.detect_face.detect_face(img, minsize, pnet, rnet, onet, threshold, factor)
returns the detected faces as rows in a numpy array.
In align_dataset_mtcnn.py
there is some code to figure out which face (among the detected ones) that is the most relevant, but that could shouldn't be needed then.
Hello) Is it possible for MTCNN to return all bounding boxes with all different faces on an image? If yes, what is the right way to do this?
Waining for respond. Thank you in advance)