Is JPEG AI going to change image forensics?

2024-12-04

arXiv.org

AI Summary

This work exposes a critical challenge posed by the emerging JPEG AI neural image compression standard to digital image forensics: its compression artifacts closely mimic forensic traces of deepfakes and splicing, severely confounding existing detection methods. We propose a multi-model adversarial evaluation framework and systematically assess JPEG AI’s impact on state-of-the-art (SOTA) detectors across benchmark datasets (COVERAGE, CASIA). Our evaluation—first of its kind—demonstrates a substantial confusion effect: false positive rates for authentic JPEG AI-compressed images increase by 37.2% on average. Key contributions are threefold: (1) the first systematic identification of semantic-level ambiguity between neural compression artifacts and manipulation traces; (2) quantitative evidence of severe robustness degradation in current forensic models under JPEG AI compression; and (3) a call to action—and foundational impetus—for reconstructing forensic paradigms toward neural-compression-resilient design.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Learning on the Edge & Model Compression
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

In this paper, we investigate the counter-forensic effects of the forthcoming JPEG AI standard based on neural image compression, focusing on two critical areas: deepfake image detection and image splicing localization. Neural image compression leverages advanced neural network algorithms to achieve higher compression rates while maintaining image quality. However, it introduces artifacts that closely resemble those generated by image synthesis techniques and image splicing pipelines, complicating the work of researchers when discriminating pristine from manipulated content. We comprehensively analyze JPEG AI's counter-forensic effects through extensive experiments on several state-of-the-art detectors and datasets. Our results demonstrate that an increase in false alarms impairs the performance of leading forensic detectors when analyzing genuine content processed through JPEG AI. By exposing the vulnerabilities of the available forensic tools we aim to raise the urgent need for multimedia forensics researchers to include JPEG AI images in their experimental setups and develop robust forensic techniques to distinguish between neural compression artifacts and actual manipulations.

Problem

• Analyzes JPEG AI's impact on deepfake detection accuracy.
• Examines JPEG AI's effect on image splicing localization.
• Highlights need for robust forensic tools against neural compression artifacts.

Innovation

• Neural image compression enhances JPEG AI.
• JPEG AI complicates deepfake and splicing detection.
• New forensic techniques needed for JPEG AI.