RealStats: A Rigorous Real-Only Statistical Framework for Fake Image Detection

Haim Zisman, Uri Shaham

TL;DR: Statistically rigorous, training-free, and adaptable method for detecting AI-generated images using only real image statistics and interpretable p-value aggregation.

Abstract: As generative models continue to evolve, detecting AI-generated images remains a critical challenge. While effective detection methods exist, they often lack formal interpretability and may rely on implicit assumptions about the nature of fake content, potentially limiting their robustness to distributional shifts. In this work, we introduce a rigorous, statistically grounded framework for fake image detection that focuses on producing a probability score interpretable with respect to the real-image population. Our method leverages the strengths of multiple existing detectors by combining strong training-free statistics. We compute $p$-values over a range of test statistics and aggregate them using classical statistical ensembling to assess alignment with the unified real-image distribution. This framework is generic, flexible, and training-free, making it well-suited for robust fake image detection across diverse and evolving settings.

Submission Number: 410