The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Online Program Home
Abstract Details
Activity Number:
|
222
|
Type:
|
Topic Contributed
|
Date/Time:
|
Monday, July 30, 2012 : 2:00 PM to 3:50 PM
|
Sponsor:
|
IMS
|
Abstract - #305959 |
Title:
|
Hidden Markov Trees and Fisher Information Distance for Artist Identification and Dating in Impressionist Paintings
|
Author(s):
|
Shannon Hughes*+
|
Companies:
|
University of Colorado at Boulder
|
Address:
|
Dept. of Electrical Engineering, Boulder, CO, 80309, United States
|
Keywords:
|
art ;
stylometry ;
Fisher information distance ;
Hidden Markov trees ;
classification
|
Abstract:
|
Visual stylometry of art proposes to apply statistical tools to high-resolution digital images of artwork to produce a quantitative description of each work's style or of the ``stylistic distance'' between works. The premise is that an artist's unique habitual movements while painting leave behind characteristic measurable stylistic features in the brushwork. Such quantitative descriptions of style can then aid art scholars in answering open art historical questions, including those of the work's authorship and date of creation. Here, we propose new methods for visual stylometry. We use the background of each painting for analysis, hypothesizing that only this bears the signature of the artist's habitual movements. Then, for each brushwork sample, we estimate the parameters of a wavelet-Hidden-Markov-Tree (WHMT) texture model. Finally, we measure Fisher information distance between the resulting WHMT parameter distributions as a metric of stylistic distance between paintings. Tests on two datasets consisting of over 100 impressionist paintings show that our stylistic distance metric tends to cluster the paintings according to author and, within an author, according to time period.
|
The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.
Back to the full JSM 2012 program
|
2012 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.