4 ms·
We do use facial recognition to detect the faces in the photo, and then integrate with the school's student management system to work out what permissions or me
by cetra3 7y ago
We do use facial recognition to detect the faces in the photo, and then integrate with the school's student management system to work out what permissions or media consent has been granted for each student.
From there you can identify whether the photo overall is publishable, and if not, what students aren't. You can then crop/edit the photo to remove students from then on.
All of this happens in an on-premise VM, rather than calling out to a cloud service. Obviously you can run up an instance in AWS, etc.. if you want to cloud host, but that is the school's volition.
- vageli 7y ago> We do use facial recognition to detect the faces in the photo, and then integrate with the school's student management system to work out what permissions or media consent has been granted for each student. > From there you can identify whether the photo overall is publishable, and if not, what students aren't. You can then crop/edit the photo to remove students from then on. > All of this happens in an on-premise VM, rather than calling out to a cloud service. Obviously you can run up an instance in AWS, etc.. if you want to cloud host, but that is the school's volition. But if the parents don't sign the waiver allowing their kids photo to be used, how do you process the photo to determine if any kid in it has not had the waiver signed? Isn't the act of processing the image a violation of the agreement (which is effective since the waiver was not signed)?