Well, what are you going to use this technique for? What’s the use case for generating a high-res photo of a face from a low-res photo of a face? It’s hard to think of any use case with real-world consequences that wouldn’t be a bad idea regardless of the training set.
> We have noticed a lot of concern that PULSE will be used to identify individuals whose faces have been blurred out. We want to emphasize that this is impossible - PULSE makes imaginary faces of people who do not exist, which should not be confused for real people. It will not help identify or reconstruct the original image.
This is just another way to generate realistic faces that match a pattern. The "bad use cases" I can imagine involve unmasking real people, which is not what the model does.
My org allows profpics for email or messaging or stuff like that. Some people have old, tiny profpics that are pixellated at various sizes. This technique could help the pictures look nicer.
