file-metadata should be enhanced on the face detection part. We need to integrate head pose estimation and face recognition. E.g use size of face (ration compared to picture size - kind of coverage) and orientation (head pose) for portrait vs. face profile detection.
See also:
- Perspective 3-Point (P3P) Algorithm: http://rpg.ifi.uzh.ch/software_datasets.html
- http://rpg.ifi.uzh.ch/software/p3p_code_final.zip
- http://rpg.ifi.uzh.ch/docs/CVPR11_kneip.pdf
- face recognition (e.g. like facebook) as well as age and gender - needs kind of DB (e.g. commons): T135836#2314683, T135836#2314835
- facial landmarks: T135836#2314683
Possible Categories to put for the bot:
Details:
Primary mentor:
Co-mentor:
Other mentors: (optional, Phabricator username)
Skills: (python and computer vision)
Estimated project time for a senior contributor: (2-4 weeks)