We construct a segmentation scheme that combines top-down with bottom-up
processing. In the proposed scheme, segmentation and recognition
are intertwined rather than proceeding in a serial manner. The top-down
part applies stored knowledge about object shapes acquired through
learning, whereas the bottom-up part creates a hierarchy of segmented
regions based on uniformity criteria. Beginning with unsegmented
training examples of class and non-class images, the algorithm constructs
a bank of class-specific fragments and determines their figure-ground
segmentation. This bank is then used to segment novel images in a
top-down manner: the fragments are first used to recognize images
containing class objects, and then to create a complete cover that
best approximates these objects. The resulting segmentation is then
integrated with bottom-up multi-scale grouping to better delineate
the object boundaries. Our experiments, applied to a large set of
four classes (horses, pedestrians, cars, faces), demonstrate segmentation
results that surpass those achieved by previous top-down or bottom-up
schemes. The main novel aspects of this work are the fragment learning
phase, which efficiently learns the figure-ground labeling of segmentation
fragments, even in training sets with high object and background
variability; combining the top-down segmentation with bottom-up criteria
to draw on their relative merits; and the use of segmentation to
improve recognition.
%0 Journal Article
%1 Borenstein2008
%A Borenstein, Eran
%A Ullman, Shimon
%D 2008
%J IEEE transactions on pattern analysis and machine intelligence
%K Algorithms,Artificial Automated,Pattern Automated: Computer-Assisted,Image Computer-Assisted: Enhancement,Image Enhancement: Intelligence,Automated,Automated: Interpretation, Recognition, Results,Sensitivity Specificity,Subtraction Technique and methods,Computer-Assisted,Computer-Assisted: methods,Image methods,Pattern methods,Reproducibility of
%N 12
%P 2109--25
%R 10.1109/TPAMI.2007.70840
%T Combined top-down/bottom-up segmentation.
%U http://ieeexplore.ieee.org/xpl/freeabs\_all.jsp?arnumber=4408584
%V 30
%X We construct a segmentation scheme that combines top-down with bottom-up
processing. In the proposed scheme, segmentation and recognition
are intertwined rather than proceeding in a serial manner. The top-down
part applies stored knowledge about object shapes acquired through
learning, whereas the bottom-up part creates a hierarchy of segmented
regions based on uniformity criteria. Beginning with unsegmented
training examples of class and non-class images, the algorithm constructs
a bank of class-specific fragments and determines their figure-ground
segmentation. This bank is then used to segment novel images in a
top-down manner: the fragments are first used to recognize images
containing class objects, and then to create a complete cover that
best approximates these objects. The resulting segmentation is then
integrated with bottom-up multi-scale grouping to better delineate
the object boundaries. Our experiments, applied to a large set of
four classes (horses, pedestrians, cars, faces), demonstrate segmentation
results that surpass those achieved by previous top-down or bottom-up
schemes. The main novel aspects of this work are the fragment learning
phase, which efficiently learns the figure-ground labeling of segmentation
fragments, even in training sets with high object and background
variability; combining the top-down segmentation with bottom-up criteria
to draw on their relative merits; and the use of segmentation to
improve recognition.
@article{Borenstein2008,
abstract = {We construct a segmentation scheme that combines top-down with bottom-up
processing. In the proposed scheme, segmentation and recognition
are intertwined rather than proceeding in a serial manner. The top-down
part applies stored knowledge about object shapes acquired through
learning, whereas the bottom-up part creates a hierarchy of segmented
regions based on uniformity criteria. Beginning with unsegmented
training examples of class and non-class images, the algorithm constructs
a bank of class-specific fragments and determines their figure-ground
segmentation. This bank is then used to segment novel images in a
top-down manner: the fragments are first used to recognize images
containing class objects, and then to create a complete cover that
best approximates these objects. The resulting segmentation is then
integrated with bottom-up multi-scale grouping to better delineate
the object boundaries. Our experiments, applied to a large set of
four classes (horses, pedestrians, cars, faces), demonstrate segmentation
results that surpass those achieved by previous top-down or bottom-up
schemes. The main novel aspects of this work are the fragment learning
phase, which efficiently learns the figure-ground labeling of segmentation
fragments, even in training sets with high object and background
variability; combining the top-down segmentation with bottom-up criteria
to draw on their relative merits; and the use of segmentation to
improve recognition.},
added-at = {2013-09-29T14:16:50.000+0200},
author = {Borenstein, Eran and Ullman, Shimon},
biburl = {https://www.bibsonomy.org/bibtex/2d8814997ca9190ce7bba5f0dd93241b6/guillem.palou},
doi = {10.1109/TPAMI.2007.70840},
interhash = {5bae61723abc32d1263a5c2925f05502},
intrahash = {d8814997ca9190ce7bba5f0dd93241b6},
issn = {0162-8828},
journal = {IEEE transactions on pattern analysis and machine intelligence},
keywords = {Algorithms,Artificial Automated,Pattern Automated: Computer-Assisted,Image Computer-Assisted: Enhancement,Image Enhancement: Intelligence,Automated,Automated: Interpretation, Recognition, Results,Sensitivity Specificity,Subtraction Technique and methods,Computer-Assisted,Computer-Assisted: methods,Image methods,Pattern methods,Reproducibility of},
number = 12,
pages = {2109--25},
pmid = {18988946},
timestamp = {2013-09-29T14:16:50.000+0200},
title = {{Combined top-down/bottom-up segmentation.}},
url = {http://ieeexplore.ieee.org/xpl/freeabs\_all.jsp?arnumber=4408584},
volume = 30,
year = 2008
}