Looks TooGoodToBeTrue: AnInformation-TheoreticAnalysisofHallucinations inGenerativeRestorationModels

The pursuit of high perceptual quality in image restoration has driven the development of revolutionary generative models, capable of producing results often visually indistinguishable from real data. However, as their perceptual quality continues toimprove, these models also exhibit agrowing tendencytogenerate hallucinations -realistic-looking details that do not exist in the ground truth images. Hallucinations in these models create uncertainty about their reliability, raising major concerns about their practical application.