Towards Universal Fake Image Detectors that Generalize Across Generative Models

arXiv:2302.10174v2 Announce Type: replace-cross
Abstract: With generative models proliferating at a rapid rate, there is a growing need for general purpose fake image detectors. In this work, we first show that the existing paradigm, which consists of training a deep network for real-vs-fake classification, fails to detect fake images from newer breeds of generative models when trained to detect GAN fake images. Upon analysis, we find that the resulting classifier is asymmetrically tuned to detect patterns that make an image …

abstract arxiv classification cs.cv cs.lg detectors fake general generative generative models image images network paradigm rate show training type universal work