Color Compatibility From Large Datasets
Peter O'Donovan1 Aseem Agarwala2 Aaron Hertzmann1 |
1University of Toronto 2Adobe Systems, Inc. |
Abstract
This paper studies color compatibility theories using large datasets, and develops new tools for choosing colors. There are three parts to this work. First, using on-line datasets, we test new and existing theories of human color preferences. For example, we test whether certain hues or hue templates may be preferred by viewers. Second, we learn quantitative models that score the quality of a set of five colors, called a color theme. Such models can be used to rate the quality of a new color theme. Third, we demonstrate simple prototypes that apply a learned model to tasks in color design, including improving existing themes and extracting themes from images.
Paper
Peter O'Donovan, Aseem Agarwala, Aaron Hertzmann. Color Compatibility From Large Datasets. ACM Transactions on Graphics, 2011, 30, 43, (Proc. SIGGRAPH). BibTexSupplemental Materials
Further Experiments: supplemental.pdf
Datasets and Code for Training/Testing a Compatibility Model: colorCode.zip
Data and code are released under the Creative Commons BY-NC-SA license. If you use this data or code for a publication, we ask that you cite the above paper.
LASSO Model Weights: weights.csv
SIGGRAPH Slides: slides.pptx
Acknowledgements
Thanks to Koji Yatani for translating Matsuda [1995]. We thank Maneesh Agrawala, Dan Goldman, and Sylvain Paris for helpful discussions. This research is supported in part by Adobe, NSERC,and CIFAR.We would also like to thank the Kuler team, COLOURLovers, Dimitri Boisdet for the Threatening Sky photo above, Stephen Zacharias for the Marinos Ices Mixture photo, and Yv-Chan for the Phoenix graphic design.