Multi-axis Analysis of Image Manipulation Localization

TL;DR

This paper introduces AUDITS, a large benchmark dataset for analyzing the robustness of image manipulation detection methods across various domains, manipulation types, and sizes, especially under domain shifts.

Contribution

The paper presents AUDITS, a comprehensive dataset with over 530K images for studying image manipulation detection across multiple axes and domain shifts, supporting future research.

Findings

Existing detection methods' robustness varies across domain shifts.

AUDITS enables analysis of manipulation detection across diverse types and sizes.

Benchmark results highlight gaps in current methods' generalizability.

Abstract

Advanced image editing software enables easy creation of highly convincing image manipulations, which has been made even more accessible in recent years due to advances in generative AI. Manipulated images, while often harmless, could spread misinformation, create false narratives, and influence people's opinions on important issues. Despite this growing threat, there is limited research on detecting advanced manipulations across different visual domains. Thus, we introduce Analysis Under Domain-shifts, qualIty, Type, and Size (AUDITS), a comprehensive benchmark designed for studying axes of analysis in image manipulation detection. AUDITS comprises over 530K images from two distinct sources (user and news photos). We curate our dataset to support analysis across multiple axes using recent diffusion-based inpaintings, spanning a diverse range of manipulation types and sizes. We conduct experiments under different types of domain shift to evaluate robustness of existing image manipulation detection methods. Our goal is to drive further research in this area by offering new insights that would help develop more reliable and generalizable image manipulation detection methods.