Characterizing Photorealism and Artifacts in Diffusion Model-Generated Images

要旨

Diffusion model-generated images can appear indistinguishable from authentic photographs, but these images often contain artifacts and implausibilities that reveal their AI-generated provenance. Given the challenge to public trust in media posed by photorealistic AI-generated images, we conducted a large-scale experiment measuring human detection accuracy on 450 diffusion-model generated images and 149 real images. Based on collecting 749,828 observations and 34,675 comments from 50,444 participants, we find that scene complexity of an image, artifact types within an image, display time of an image, and human curation of AI-generated images all play significant roles in how accurately people distinguish real from AI-generated images. Additionally, we propose a taxonomy characterizing artifacts often appearing in images generated by diffusion models. Our empirical observations and taxonomy offer nuanced insights into the capabilities and limitations of diffusion models to generate photorealistic images in 2024.

著者
Negar Kamali
Northwestern University, Evanston, Illinois, United States
Karyn Nakamura
Northwestern University, Evanston, Illinois, United States
Aakriti Kumar
Northwestern University, Evanston, Illinois, United States
Angelos Chatzimparmpas
Utrecht University, Utrecht, Netherlands
Jessica Hullman
Northwestern University, Evanston, Illinois, United States
Matthew Groh
Northwestern, Evanston, Illinois, United States
DOI

10.1145/3706598.3713962

論文URL

https://dl.acm.org/doi/10.1145/3706598.3713962

動画

会議: CHI 2025

The ACM CHI Conference on Human Factors in Computing Systems (https://chi2025.acm.org/)

セッション: Optimization with/for AI

G318+G319
7 件の発表
2025-04-30 23:10:00
2025-05-01 00:40:00
日本語まとめ
読み込み中…