There’s a lot of talk of undercover cops intentionally starting violence in otherwise peaceful protests. The evidence, primarily, are images like this.
https://images.app.goo.gl/4n3o2EXwFzMQfsKq6
It looks pretty convincing, but I’d like something more solid. Does anyone know of a model that can detect with a high level of certainty if the “mask” area of two photos represents the same person?
The model you are looking for is probably a Siamese Neural Networks.
It's very used for Biometrics. This kind of model learns to extract relevant and features (For example, facial metrics) that are invariant on several conditions (that won't change on light, angle, or even occlusions) presented on the dataset.
But keep in mind no model is an Scy-Fi AI, like FBI movies enhancing the image 1000x. There is natural limitation, so maybe not even the best AI possible could give you a satisfactory answer.
