[Python] Find most similar person based on faces where each person has a different amount of...

I am stuck on the following problem: I have a database of persons p_i, where each person has k_p_i many face-pictures with face-encodings (dependent on the person).

Now lets say I input any picture of a human face and I want to match it to a person. How do I weigh the differences in amount of pictures? If every person in the database has only one picture, i would just take the person whose picture has the best similarity. If every person has the same amount of pictures I would just use the average. In my case however I do want to prioritize people with more images, simply because I have more proof to believe that they are the same person.

Is there any canonical algorithm/best practice/recommendations you have?

Continue reading...