[Python] How to train an eye verification (not recognition) model in PyTorch?

I want to know how we train one-to-one image verification models. It’s where a model takes two images and verifies that they are both the same.

I have googled online and I can only find answers about recognition (one-to-many).

Can anyone please explain how to create such a model in either text or code? I will be grateful.

Edit: For clarification, by the same I mean that the eyes are the same, meaning they return to the same person. It’s a verification model.

Continue reading...