analiza aiqbench
The introduction of SynCred-Bench reveals a critical vulnerability in current AI safety systems, where even commercial APIs struggle to detect synthetic credibility in images containing realistic text and layouts. Under strict false-positive constraints, existing models and human annotators fail to identify these threats with high reliability, highlighting the urgent need for detectors that can reason beyond superficial visual cues. This benchmark establishes synthetic credibility as a severe, underexplored challenge that demands immediate attention from researchers developing robust misinformation defenses.
SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation
arXiv:2606.03348v1 Announce Type: cross Abstract: Recent generative models can now produce visual artifacts with realistic embedded text and layouts, creating a new misinformation threat: synthetic credibility. We introduce SYNCRED-Bench, a benchmark of 600 AI-generated misinformation images balanced across six credible-form categories and seven fine-grained circulation styles, together with FP450, a real-image negative set for measuring false positives. Extensive evaluation shows that existing systems remain unreliable: under a 5% false-positive-rate constraint, 15 MLLMs achieve only 10.5% true positive rate (TPR), open-source AIGC detectors achieve less than 5%, and commercial APIs reach 57.6%. Human annotators also struggled to identify synthetic credibility, reaching only 63% TPR. These findings establish synthetic credibility as a severe and underexplored visual misinformation challenge, and provide a benchmark for developing detectors that reason beyond superficial credibility cues.
Abstract:Recent generative models can now produce visual artifacts with realistic embedded text and layouts, creating a new misinformation threat: synthetic credibility. We introduce SYNCRED-Bench, a benchmark of 600 AI-generated misinformation images balanced across six credible-form categories and seven fine-grained circulation styles, together with FP450, a real-image negative set for measuring false positives. Extensive evaluation shows that existing systems remain unreliable: under a 5% false-positive-rate constraint, 15 MLLMs achieve only 10.5% true positive rate (TPR), open-source AIGC detectors achieve less than 5%, and commercial APIs reach 57.6%. Human annotators also struggled to identify synthetic credibility, reaching only 63% TPR. These findings establish synthetic credibility as a severe and underexplored visual misinformation challenge, and provide a benchmark for developing detectors that reason beyond superficial credibility cues.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
arXiv:2606.03348 [cs.CV]
(or arXiv:2606.03348v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2606.03348 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 224 cuvinte