Epistemic Uncertainty for Generated Image Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Nie, Jun, Zhang, Yonggang, Liu, Tongliang, Cheung, Yiu-ming, Han, Bo, Tian, Xinmei
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909881157550080
author Nie, Jun
Zhang, Yonggang
Liu, Tongliang
Cheung, Yiu-ming
Han, Bo
Tian, Xinmei
author_facet Nie, Jun
Zhang, Yonggang
Liu, Tongliang
Cheung, Yiu-ming
Han, Bo
Tian, Xinmei
contents We introduce a novel framework for AI-generated image detection through epistemic uncertainty, aiming to address critical security concerns in the era of generative models. Our key insight stems from the observation that distributional discrepancies between training and testing data manifest distinctively in the epistemic uncertainty space of machine learning models. In this context, the distribution shift between natural and generated images leads to elevated epistemic uncertainty in models trained on natural images when evaluating generated ones. Hence, we exploit this phenomenon by using epistemic uncertainty as a proxy for detecting generated images. This converts the challenge of generated image detection into the problem of uncertainty estimation, underscoring the generalization performance of the model used for uncertainty estimation. Fortunately, advanced large-scale vision models pre-trained on extensive natural images have shown excellent generalization performance for various scenarios. Thus, we utilize these pre-trained models to estimate the epistemic uncertainty of images and flag those with high uncertainty as generated. Extensive experiments demonstrate the efficacy of our method. Code is available at https://github.com/tmlr-group/WePe.
format Preprint
id arxiv_https___arxiv_org_abs_2412_05897
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Epistemic Uncertainty for Generated Image Detection
Nie, Jun
Zhang, Yonggang
Liu, Tongliang
Cheung, Yiu-ming
Han, Bo
Tian, Xinmei
Computer Vision and Pattern Recognition
Artificial Intelligence
We introduce a novel framework for AI-generated image detection through epistemic uncertainty, aiming to address critical security concerns in the era of generative models. Our key insight stems from the observation that distributional discrepancies between training and testing data manifest distinctively in the epistemic uncertainty space of machine learning models. In this context, the distribution shift between natural and generated images leads to elevated epistemic uncertainty in models trained on natural images when evaluating generated ones. Hence, we exploit this phenomenon by using epistemic uncertainty as a proxy for detecting generated images. This converts the challenge of generated image detection into the problem of uncertainty estimation, underscoring the generalization performance of the model used for uncertainty estimation. Fortunately, advanced large-scale vision models pre-trained on extensive natural images have shown excellent generalization performance for various scenarios. Thus, we utilize these pre-trained models to estimate the epistemic uncertainty of images and flag those with high uncertainty as generated. Extensive experiments demonstrate the efficacy of our method. Code is available at https://github.com/tmlr-group/WePe.
title Epistemic Uncertainty for Generated Image Detection
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2412.05897