There are algorithms developed by Microsoft and others which are specifically designed to normalize images as best they can before processing in order to reduce

There are algorithms developed by Microsoft and others which are specifically designed to normalize images as best they can before processing in order to reduce false negatives due to alterations like cropping, color shifting, watermarking, etc. They then use perceptual hashing techniques to identify similar images. It's a game of cat and mouse but ultimately I think with ML, the mouse is going to lose.