Object Categorization

Research output: Contribution to journalArticleResearchpeer-review

Abstract

This article presents foundations, original research and trends in the
field of object categorization by computer vision methods. The research
goals in object categorization are to detect objects in images and to
determine the object’s categories. Categorization aims for the recognition
of generic classes of objects, and thus has also been termed
‘generic object recognition’. This is in contrast to the recognition of
specific, individual objects. While humans are usually better in generic
than in specific recognition, categorization is much harder to achieve
for today’s computer architectures and algorithms. Major problems are
related to the concept of a ‘visual category’, where a successful recognition
algorithm has to manage large intra-class variabilities versus
sometimes marginal inter-class differences. It turns out that several
techniques which are useful for specific recognition can also be adapted
to categorization, but there are also a number of recent developments
in learning, representation and detection that are especially tailored to
categorization.
Recent results have established various categorization methods that
are based on local salient structures in the images. Some of these methods
use just a ‘bag of keypoints’ model. Others include a certain amount
of geometric modeling of 2D spatial relations between parts, or ‘constellations’
of parts. There is now a certain maturity in these approaches
and they achieve excellent recognition results on rather complex image
databases. Further work focused on the description of shape and object
contour for categorization is only just emerging. However, there remain
a number of important open questions, which also define current and
future research directions. These issues include localization abilities,
required supervision, the handling of many categories, online and incremental
learning, and the use of a ‘visual alphabet’, to name a few. These
aspects are illustrated by the discussion of several current approaches,
including our own patch-based system and our boundary fragmentmodel.
The article closes with a summary and a discussion of promising
future research directions.
Original languageEnglish
Pages (from-to)255-353
JournalFoundations and trends in computer graphics and vision
Volume1
Issue number4
DOIs
Publication statusPublished - 2006

Fingerprint

Computer architecture
Object recognition
Computer vision

Treatment code (Nähere Zuordnung)

  • Review

Cite this

Object Categorization. / Pinz, Axel.

In: Foundations and trends in computer graphics and vision, Vol. 1, No. 4, 2006, p. 255-353.

Research output: Contribution to journalArticleResearchpeer-review

@article{c5e9efd69d9b467bb50e7d0610250935,
title = "Object Categorization",
abstract = "This article presents foundations, original research and trends in thefield of object categorization by computer vision methods. The researchgoals in object categorization are to detect objects in images and todetermine the object’s categories. Categorization aims for the recognitionof generic classes of objects, and thus has also been termed‘generic object recognition’. This is in contrast to the recognition ofspecific, individual objects. While humans are usually better in genericthan in specific recognition, categorization is much harder to achievefor today’s computer architectures and algorithms. Major problems arerelated to the concept of a ‘visual category’, where a successful recognitionalgorithm has to manage large intra-class variabilities versussometimes marginal inter-class differences. It turns out that severaltechniques which are useful for specific recognition can also be adaptedto categorization, but there are also a number of recent developmentsin learning, representation and detection that are especially tailored tocategorization.Recent results have established various categorization methods thatare based on local salient structures in the images. Some of these methodsuse just a ‘bag of keypoints’ model. Others include a certain amountof geometric modeling of 2D spatial relations between parts, or ‘constellations’of parts. There is now a certain maturity in these approachesand they achieve excellent recognition results on rather complex imagedatabases. Further work focused on the description of shape and objectcontour for categorization is only just emerging. However, there remaina number of important open questions, which also define current andfuture research directions. These issues include localization abilities,required supervision, the handling of many categories, online and incrementallearning, and the use of a ‘visual alphabet’, to name a few. Theseaspects are illustrated by the discussion of several current approaches,including our own patch-based system and our boundary fragmentmodel.The article closes with a summary and a discussion of promisingfuture research directions.",
author = "Axel Pinz",
year = "2006",
doi = "10.1561/0600000003",
language = "English",
volume = "1",
pages = "255--353",
journal = "Foundations and trends in computer graphics and vision",
issn = "1572-2740",
publisher = "Now Publishers Inc",
number = "4",

}

TY - JOUR

T1 - Object Categorization

AU - Pinz, Axel

PY - 2006

Y1 - 2006

N2 - This article presents foundations, original research and trends in thefield of object categorization by computer vision methods. The researchgoals in object categorization are to detect objects in images and todetermine the object’s categories. Categorization aims for the recognitionof generic classes of objects, and thus has also been termed‘generic object recognition’. This is in contrast to the recognition ofspecific, individual objects. While humans are usually better in genericthan in specific recognition, categorization is much harder to achievefor today’s computer architectures and algorithms. Major problems arerelated to the concept of a ‘visual category’, where a successful recognitionalgorithm has to manage large intra-class variabilities versussometimes marginal inter-class differences. It turns out that severaltechniques which are useful for specific recognition can also be adaptedto categorization, but there are also a number of recent developmentsin learning, representation and detection that are especially tailored tocategorization.Recent results have established various categorization methods thatare based on local salient structures in the images. Some of these methodsuse just a ‘bag of keypoints’ model. Others include a certain amountof geometric modeling of 2D spatial relations between parts, or ‘constellations’of parts. There is now a certain maturity in these approachesand they achieve excellent recognition results on rather complex imagedatabases. Further work focused on the description of shape and objectcontour for categorization is only just emerging. However, there remaina number of important open questions, which also define current andfuture research directions. These issues include localization abilities,required supervision, the handling of many categories, online and incrementallearning, and the use of a ‘visual alphabet’, to name a few. Theseaspects are illustrated by the discussion of several current approaches,including our own patch-based system and our boundary fragmentmodel.The article closes with a summary and a discussion of promisingfuture research directions.

AB - This article presents foundations, original research and trends in thefield of object categorization by computer vision methods. The researchgoals in object categorization are to detect objects in images and todetermine the object’s categories. Categorization aims for the recognitionof generic classes of objects, and thus has also been termed‘generic object recognition’. This is in contrast to the recognition ofspecific, individual objects. While humans are usually better in genericthan in specific recognition, categorization is much harder to achievefor today’s computer architectures and algorithms. Major problems arerelated to the concept of a ‘visual category’, where a successful recognitionalgorithm has to manage large intra-class variabilities versussometimes marginal inter-class differences. It turns out that severaltechniques which are useful for specific recognition can also be adaptedto categorization, but there are also a number of recent developmentsin learning, representation and detection that are especially tailored tocategorization.Recent results have established various categorization methods thatare based on local salient structures in the images. Some of these methodsuse just a ‘bag of keypoints’ model. Others include a certain amountof geometric modeling of 2D spatial relations between parts, or ‘constellations’of parts. There is now a certain maturity in these approachesand they achieve excellent recognition results on rather complex imagedatabases. Further work focused on the description of shape and objectcontour for categorization is only just emerging. However, there remaina number of important open questions, which also define current andfuture research directions. These issues include localization abilities,required supervision, the handling of many categories, online and incrementallearning, and the use of a ‘visual alphabet’, to name a few. Theseaspects are illustrated by the discussion of several current approaches,including our own patch-based system and our boundary fragmentmodel.The article closes with a summary and a discussion of promisingfuture research directions.

UR - http://www.nowpublishers.com

U2 - 10.1561/0600000003

DO - 10.1561/0600000003

M3 - Article

VL - 1

SP - 255

EP - 353

JO - Foundations and trends in computer graphics and vision

JF - Foundations and trends in computer graphics and vision

SN - 1572-2740

IS - 4

ER -