Article citationsMore>>
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., Krueger, G., & Sutskever, I. (2021). Learning Transferable Visual Models from Natural Language Supervision. In M. Meila, & T. Zhang (Eds.), Proceedings of the 38th International Conference on Machine Learning (Vol. 139, pp. 8748-8763). PMLR.
https://proceedings.mlr.press/v139/radford21a.html
has been cited by the following article:
-
TITLE:
A Computational Comparison of Human Artworks and Manually Curated DALL·E 2 Images with Matched Styles
AUTHORS:
Yakun Gu, Yiying Zhang, Tianling Huang
KEYWORDS:
Generative Artificial Intelligence, Visual Art, DALL·E 2, Image Statistics, Paired Comparison, Curation
JOURNAL NAME:
Art and Design Review,
Vol.14 No.4,
November
19,
2026
ABSTRACT: This exploratory secondary study compares 50 human artworks with 50 independently generated, manually style-matched DALL·E 2 images from an official stimulus archive. Thirteen color, grayscale, local-transition, and frequency features were computed after common RGB decoding and resizing. Paired estimation used 100,000 BCa bootstrap resamples and Holm correction across 13 tests. The AI images showed higher mean green intensity (Δ = 0.0537), blue-channel dispersion (Δ = 0.0448), and grayscale entropy (Δ = 0.241 bits), and lower high-frequency energy (Δ = −0.0305). Blue dispersion and entropy retained corrected support under equal artist weighting. Green and frequency estimates barely changed, but their adjusted tests were less conclusive with 28 artist means. Matching advantages in grayscale cosine and RGB histograms were attenuated but remained supported under genre-restricted permutations; subgroup evidence was less precise for still life. A ten-pair external variation sample yielded no corrected effects. Common archived JPEG encoding did not resolve unknown upstream color and compression histories. These findings describe selected visual differences in a curated corpus and do not establish transformations of individual artworks or explain aesthetic preference.