Open arpsyapathy opened 5 years ago
what could work :
Start with the first image
find all faces
calculate face encodings for all faces and store in list "known faces" (the list could now have 1-4 entries), if it only contains 1 face, then you could already stop. you found yourself.
loop over all remaining images -- compare_faces and evaluate how often you found each face (e.g. you could create a dict and increment a counter)
the face you found most often, is you.
repeat once you know which face is you.
have a look at https://github.com/ageitgey/face_recognition/blob/master/docs/usage.rst and the examples
Or find the image with only one face and calculate the face encoding...?
To add to the above mentioned way, if you have the encodings of your face in your database (eg. Excel Sheet) then you could try matching the 4 faces in 1 photo against your database.
If it matches with your preprocessed encodings them you are good to go.
Just iterate over 5 images though to get all faces.
Hello!
I have 5 photos with me and my friends. Each photo have 1-4 faces. But each photo have my face.
How get my face encoding from this 5 photos without input data?
For example: i say "Hey script, give me my face encoding". Script analyzes 5 photos. Finds the face is present on most photo and gives it to me.
This is possible? Thank you