I am using LBPH model yml to same images after train, it is overwriting.. someone to guide me over there.. do i have to use one yml file separete yml file for each user who register his/her face.

edit

Thank you berak. do i have to store one yml file or Need to have multiple of them for every user who register the face?

cmon, one file. you read it in on start, update with more persons, save again.

yea its actually working thank you.. now the problem is when i am authenticating it is not giving the right prediction with the LBPH algorithm

now the problem is when i am authenticating

if you really want to authenticate a single user, then it's the wrong algorithm (it does identification, not authentification)

yea ok i meant prediction, now am in a possition to save one yml file and append new faces, I am saving with the labels. My challange is when predicting I am passing the image to be predicted, label and confidence. In this lbph algorithm, the label passed is -1 now confused if this label I should pass the label i set during training? Kindly assist me there.