arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.03214cs.CVcs.AIcs.LG
收藏

有机还是扩散:我们能否区分人类艺术与AI生成图像?

Organic or Diffused: Can We Distinguish Human Art from AI-generated Images?

University of Chicago(芝加哥大学);Concept Artist(概念艺术家)

Anna Yoo Jeong Ha, Josephine Passananti, Ronik Bhaskar, Shawn Shan, Reid Southen, Haitao Zheng, Ben Y. Zhao

AI总结:

本文构建了涵盖7种风格人类艺术与5个生成模型图像的数据集,系统评估5种自动检测器和3类人类群体,发现人类与自动检测器结合可提供最佳的准确性与鲁棒性。

摘要:

生成式AI图像的出现彻底颠覆了艺术世界。区分AI生成图像与人类艺术是一个具有挑战性的问题,其影响随时间推移日益增大。未能解决这一问题将使不良行为者得以欺诈那些为人类艺术支付高价的个人,以及那些明令禁止AI图像的公司。对于希望确立版权的内容所有者,以及有意筛选训练数据以避免潜在模型崩溃的模型训练者而言,这一问题也至关重要。区分人类艺术与AI图像存在多种不同方法,包括通过监督学习训练的分类器、针对扩散模型的研究工具,以及由专业艺术家利用其艺术技法知识进行的识别。在本文中,我们试图了解这些方法在面对当今现代生成模型时,在良性和对抗性环境下的表现如何。我们收集了7种风格的真实人类艺术,从5个生成模型生成匹配图像,并应用了8种检测器(5种自动检测器和3个不同的人类群体,包括180名众包工作者、4000多名专业艺术家以及13名在检测AI方面经验丰富的专家艺术家)。Hive和专家艺术家都表现非常出色,但犯错方式不同(Hive在对抗性扰动下较弱,而专家艺术家则产生更高的假阳性)。我们相信,随着模型不断演进,这些弱点仍将存在,并利用我们的数据证明,人类与自动检测器相结合的团队能够提供准确性与鲁棒性的最佳组合。

英文摘要:

The advent of generative AI images has completely disrupted the art world. Distinguishing AI generated images from human art is a challenging problem whose impact is growing over time. A failure to address this problem allows bad actors to defraud individuals paying a premium for human art and companies whose stated policies forbid AI imagery. It is also critical for content owners to establish copyright, and for model trainers interested in curating training data in order to avoid potential model collapse. There are several different approaches to distinguishing human art from AI images, including classifiers trained by supervised learning, research tools targeting diffusion models, and identification by professional artists using their knowledge of artistic techniques. In this paper, we seek to understand how well these approaches can perform against today's modern generative models in both benign and adversarial settings. We curate real human art across 7 styles, generate matching images from 5 generative models, and apply 8 detectors (5 automated detectors and 3 different human groups including 180 crowdworkers, 4000+ professional artists, and 13 expert artists experienced at detecting AI). Both Hive and expert artists do very well, but make mistakes in different ways (Hive is weaker against adversarial perturbations while Expert artists produce higher false positives). We believe these weaknesses will remain as models continue to evolve, and use our data to demonstrate why a combined team of human and automated detectors provides the best combination of accuracy and robustness.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑