Characterizing Photorealism and Artifacts in Diffusion Model-Generated Images
This addresses the challenge to public trust in media posed by photorealistic AI-generated images, providing empirical insights into detection capabilities.
The study measured human ability to distinguish diffusion model-generated images from real photographs, finding that factors like scene complexity, artifact types, display time, and human curation significantly affect detection accuracy, with data from 749,828 observations and 34,675 comments from 50,444 participants.
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.