This is a site to show the result of the survey conducted for quantitative evavluation in the paper “Proxy Learning of Visual Concepts of Fine Art Paintings from Styles through Language Models” (AAAI-22).
Using the three ordinal scales of (1) relevant, (2) somewhat relevant, and (3) irrelevant, the paintings are measured how they are related to each of the visual concepts.
This data is compiled by art history and Halsey Institute of Contemporary Art faculty at College of Charleston.
Downable files are available in GitHub .
Questions or Requests to Diana Kim: firstname.lastname@example.org