Grounding and Explaining Visual Evidence for AI-Generated Image Detection in Human-Centric Scenes

2026-08-03

AI Summary

This work addresses the limitations of existing AI-generated image detection methods, which often lack reliable and consistent interpretable visual evidence—particularly in human-centric scenes—and suffer from insufficient benchmark coverage. To this end, the authors introduce HAVE, a large-scale, fine-grained annotated dataset of real and synthetic human-centric images, and propose PAVE, a novel framework that jointly models authenticity discrimination, visual evidence localization, and region-aligned explanation generation. PAVE incorporates a judge-guided alignment reward mechanism to evaluate consistency between localized regions and their textual explanations, along with a perceptual regularization strategy that enhances reliance on visual input by contrasting token-level predictions from original and masked images. Experiments demonstrate that PAVE significantly outperforms current approaches across authenticity detection, evidence localization, and explanation quality on both HAVE and multiple external benchmarks.

Abstract

Rapid advances in image generation models call for interpretable AI-generated image detection methods that not only determine authenticity but also provide supporting visual evidence. Existing approaches may produce inconsistencies between generated explanations and localized evidence regions, undermining the reliability of explanations for authenticity decisions. Meanwhile, existing benchmarks provide limited coverage of the diverse human-centric scenes prevalent in generated imagery. To address these limitations, we investigate authenticity detection with grounded and explainable visual evidence in human-centric scenes. We present HAVE (Human-centric AI-generated Visual Evidence), a diverse human-centric dataset comprising 40K real and 39K AI-generated images from 10 recent generators, with 106K localized evidence instances across 8 evidence categories, each annotated with a bounding box and a region-aligned explanation. We further propose PAVE, a Perception-Aware Visual Evidence framework that jointly performs authenticity prediction, visual evidence grounding, and region-aligned explanation generation. PAVE employs a judge-guided alignment reward to assess region--explanation consistency and evidence validity, together with perception-aware regularization that contrasts token-level predictions between original and randomly masked images to promote reliance on visual input. Experiments on HAVE and external datasets demonstrate strong performance in authenticity detection, visual evidence grounding, and explanation quality. Code and data will be released upon publication.

Problem

• AI-generated image detection
• visual evidence grounding
• explainable AI
• human-centric scenes
• interpretability

Innovation

• AI-generated image detection
• visual evidence grounding
• explainable AI
• human-centric scenes
• perception-aware regularization