Color statistics of images created by generative AI
Article
Article Title | Color statistics of images created by generative AI |
---|---|
ERA Journal ID | 34297 |
Article Category | Article |
Authors | Wang, Yujin, Webster, Mike A. and Joyce, Daniel S. |
Journal Title | Journal of the Optical Society of America A: Optics, Image Science and Vision |
Journal Citation | 42 (5), pp. B76-B80 |
Number of Pages | 5 |
Year | 2025 |
Publisher | Optical Society of America (OSA) |
Place of Publication | United States |
ISSN | 1084-7529 |
1520-8532 | |
Digital Object Identifier (DOI) | https://doi.org/10.1364/JOSAA.545030 |
Web Address (URL) | https://opg.optica.org/josaa/abstract.cfm?uri=josaa-42-5-B76 |
Abstract | Novel digital images are increasingly created using generative AI tools. However, how well these AI images reflect the color statistics of human-generated (“real”) images is unknown. We analyzed hue, chroma, and lightness distributions for image objects and backgrounds created by three generative AI models (Open AI DALL E 2, Stability AI Dreamstudio, and Adobe Firefly) and for real images (Bing image search); N = 8400 images. We found that, although AI images had similar mean hues to real images, AI images systematically exhibited statistical anomalies, including narrower tunings for hue, chroma, and lightness, and higher correlations between their objects and backgrounds. While a conscious, perceptual appraisal may consider AI images to be high-fidelity, our results indicate that AI images are underpinned by color statistics that are unnatural for visual sensation. As generative AI increasingly dominates the creation of novel digital media, we may experience a narrowing of our color diet. |
Keywords | generative AI; color ; digital images |
Contains Sensitive Content | Does not contain sensitive content |
ANZSRC Field of Research 2020 | 520299. Biological psychology not elsewhere classified |
321204. Vision science | |
Byline Affiliations | Centre for Health Research |
School of Psychology and Wellbeing | |
University of Nevada, United States |
https://research.usq.edu.au/item/zv496/color-statistics-of-images-created-by-generative-ai
13
total views0
total downloads6
views this month0
downloads this month