Automatic Ranking of Iconic Images

Tamara Lee Berg and David Forsyth

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2007-13
January 12, 2007

http://www.eecs.berkeley.edu/Pubs/TechRpts/2007/EECS-2007-13.pdf

We define an iconic image for an object category (\eg eiffel tower) as an image with a large clearly delineated instance of the object in a characteristic aspect. We show that for a variety of objects such iconic images exist and argue that these are the images most relevant to that category. Given a large set of images noisily labeled with a common theme, say a Flickr tag, we show how to rank these images according to how well they represent a visual category. We also generate a binary segmentation for each image indicating roughly where the subject is located. The segmentation procedure is learned from data on a small set of iconic images from a few training categories and then applied to several other test categories. We rank the segmented test images according to shape and appearance similarity against a set of 5 hand-labeled images per category. We compute three rankings of the data: a random ranking of the images within the category, a ranking using similarity over the whole image, and a ranking using similarity applied only within the subject of the photograph. We then evaluate the rankings qualitatively and with a user study.


BibTeX citation:

@techreport{Berg:EECS-2007-13,
    Author = {Berg, Tamara Lee and Forsyth, David},
    Title = {Automatic Ranking of Iconic Images},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {2007},
    Month = {Jan},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/2007/EECS-2007-13.html},
    Number = {UCB/EECS-2007-13},
    Abstract = {We define an iconic image for an object category (\eg eiffel tower) as an image with a large clearly delineated instance of the object in a characteristic aspect. We show that for a variety of objects such iconic images exist and argue that these are the images most relevant to that category. Given a large set of images noisily labeled with a common theme, say a Flickr tag, we show how to rank these images according to how well they represent a visual category.  We also generate a binary segmentation for each image indicating roughly where the subject is located. The segmentation procedure is learned from data on a small set of iconic images from a few training categories and then applied to several other test categories. We rank the segmented test images according to shape and appearance similarity against a set of 5 hand-labeled images per category. We compute three rankings of the data: a random ranking of the images within the category, a ranking using similarity over the whole image, and a ranking using similarity applied only within the
subject of the photograph. We then evaluate the rankings qualitatively and with a user study.}
}

EndNote citation:

%0 Report
%A Berg, Tamara Lee
%A Forsyth, David
%T Automatic Ranking of Iconic Images
%I EECS Department, University of California, Berkeley
%D 2007
%8 January 12
%@ UCB/EECS-2007-13
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/2007/EECS-2007-13.html
%F Berg:EECS-2007-13