Face trained modal works but detects all faces and the faces it was supposed to detect

Hey all!

I have trained a TensorFlow modal with faces of people I want to detect.

While it detects the people and gives the correct label to the faces, I trained the model to detect. If I point the webcam at a face that I did not train the model with, it gives a label of one of the people I trained the model with.

I’ve tried many things to stop this, but nothing has worked.

I can share all the code and faces I am trying to detect if needed, but is there any way to stop this?

Any advice is greatly appreciated! I’m still learning TensorFlow and while I’m a little better than my pervious posts I’m still learning!

Thanks!