Ask Your Question

Use of LBPHFaceRecognizer (python)

asked 2019-04-10 02:54:13 -0500

procton gravatar image

I am testing the python version of LBPHFaceRecognizer and results seems not bad, but I noted something strange. Making the prediction, let's say:

label= face_recognizer.predict(gray)
print('prediction', label)

I see that label is a tuple with 2 values, the first is the prediction, the second I guess is the probability. Now when probability is low the prediction is fine, while when it is high, I have false positive.

Maybe this value is not the probability, so the question is, what is its real meaning ? Thanks

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2019-04-10 03:13:36 -0500

berak gravatar image

it's not the probability, but the distance (chi_sqr is used here) to the closest neighbour in the train db, so the "inverse" of a probability.

edit flag offensive delete link more


Thank you !

procton gravatar imageprocton ( 2019-04-10 03:44:04 -0500 )edit

Question Tools

1 follower


Asked: 2019-04-10 02:54:13 -0500

Seen: 262 times

Last updated: Apr 10 '19