Open andihaki opened 3 years ago
I think so. But you need to know the indices of the lips marks first.
And I would recommend using this repo instead:
Thank for replying 👍
as far as I know, from this picture we can the index for lips start from point 49-68 https://ibug.doc.ic.ac.uk/media/uploads/images/annotpics/figure_68_markup.jpg
any idea how I can [ ] extract (get only lips landmark), then [ ] convert the annotation for csv [ ] convert csv to tfrecord [ ] finally can train model for lips detection
IF any suggestion, really appreciate it @yinguobing
Hi @yinguobing ,
in this code https://github.com/yinguobing/image_utility/blob/master/extract_face_from_ibug.py
can we extract only some point (not exact 68 point). like I want to take the lips point (facelandmark keypoint) only