Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Comp Stat

Statistics, vision, and the analysis of artistic style

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

Abstract In the field of literature, there is an established set of techniques that have been successfully leveraged in the statistical analysis of literary style, most often to answer questions of authenticity and attribution. With the digitization of huge troves of art images come significant opportunities for the development of statistical techniques for the analysis of artistic style. In this article, we suggest that the progress made and statistical techniques developed in understanding visual processing as it relates to natural scenes can serve as a useful model and inspiration for visual stylometric analysis. WIREs Comput Stat 2012, 4:115–123. doi: 10.1002/wics.197 This article is categorized under: Statistical Learning and Exploratory Methods of the Data Sciences > Pattern Recognition Applications of Computational Statistics > Psychometrics Applications of Computational Statistics > Signal and Image Processing and Coding

Histogram of basis function spatial frequency bandwidths for the natural image basis (blue) and the images shown in Figure 2c and 2d (orange and green, respectively).

[ Normal View | Magnified View ]

Two sets of images from the database used in our analysis. Image (a) corresponds to the image that had the weakest similarity to the natural images, using the histogram intersection statistic on the spatial frequency and orientation bandwidth distributions. Image (b) is the art image that was maximally similar to the natural images. Note that it depicts a common natural scene (albeit in a painterly manner). Image (c) is the art image that produced the basis functions most different from the natural image basis functions, according to the histogram images between the spatial frequency bandwidth distributions only. Image (d) is a detail of the art image that was maximally similar under the same analysis (i.e., considering only spatial frequency information). Images (a)–(c) are courtesy of the Herbert F. Johnson Museum of Art, Cornell University, and image (d) is courtesy of the Metropolitan Museum of Art, New York.

[ Normal View | Magnified View ]

Two set of basis functions trained using the sparse coding model. The set (a) was trained on the natural image set used in the original work of Olshausen and Field.29 The set (b) corresponds to the art image (shown in Figure 2). This produced a set of basis functions that had minimal average histogram intersection using the distributions of spatial frequency and orientation bandwidth with respect to the basis functions (a).

[ Normal View | Magnified View ]

Related Articles

Scientific Visualization

Browse by Topic

Applications of Computational Statistics > Psychometrics
Statistical Learning and Exploratory Methods of the Data Sciences > Pattern Recognition
Applications of Computational Statistics > Signal and Image Processing and Coding

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts