Probabilistic methods for finding people

Research output: Contribution to journalArticle

Abstract

Finding people in pictures presents a particularly difficult object recognition problem. We show how to find people by finding candidate body segments, and then constructing assemblies of segments that are consistent with the constraints on the appearance of a person that result from kinematic properties. Since a reasonable model of a person requires at least nine segments, it is not possible to inspect every group, due to the huge combinatorial complexity. We propose two approaches to this problem. In one, the search can be pruned by using projected versions of a classifier that accepts groups corresponding to people. We describe an efficient projection algorithm for one popular classifier, and demonstrate that our approach can be used to determine whether images of real scenes contain people. The second approach employs a probabilistic framework, so that we can draw samples of assemblies, with probabilities proportional to their likelihood, which allows to draw human-like assemblies more often than the non-person ones. The main performance problem is in segmentation of images, but the overall results of both approaches on real images of people are encouraging.

Original languageEnglish (US)
Pages (from-to)45-68
Number of pages24
JournalInternational Journal of Computer Vision
Volume43
Issue number1
DOIs
StatePublished - Jun 1 2001
Externally publishedYes

Fingerprint

Classifiers
Object recognition
Kinematics

Keywords

  • Grouping correspondence search
  • Human detection
  • Object recognition
  • Probabilistic inference

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Probabilistic methods for finding people. / Ioffe, S.; Forsyth, David Alexander.

In: International Journal of Computer Vision, Vol. 43, No. 1, 01.06.2001, p. 45-68.

Research output: Contribution to journalArticle

@article{9cde3e8d0de04d6693a47cf7d2951474,
title = "Probabilistic methods for finding people",
abstract = "Finding people in pictures presents a particularly difficult object recognition problem. We show how to find people by finding candidate body segments, and then constructing assemblies of segments that are consistent with the constraints on the appearance of a person that result from kinematic properties. Since a reasonable model of a person requires at least nine segments, it is not possible to inspect every group, due to the huge combinatorial complexity. We propose two approaches to this problem. In one, the search can be pruned by using projected versions of a classifier that accepts groups corresponding to people. We describe an efficient projection algorithm for one popular classifier, and demonstrate that our approach can be used to determine whether images of real scenes contain people. The second approach employs a probabilistic framework, so that we can draw samples of assemblies, with probabilities proportional to their likelihood, which allows to draw human-like assemblies more often than the non-person ones. The main performance problem is in segmentation of images, but the overall results of both approaches on real images of people are encouraging.",
keywords = "Grouping correspondence search, Human detection, Object recognition, Probabilistic inference",
author = "S. Ioffe and Forsyth, {David Alexander}",
year = "2001",
month = "6",
day = "1",
doi = "10.1023/A:1011179004708",
language = "English (US)",
volume = "43",
pages = "45--68",
journal = "International Journal of Computer Vision",
issn = "0920-5691",
publisher = "Springer Netherlands",
number = "1",

}

TY - JOUR

T1 - Probabilistic methods for finding people

AU - Ioffe, S.

AU - Forsyth, David Alexander

PY - 2001/6/1

Y1 - 2001/6/1

N2 - Finding people in pictures presents a particularly difficult object recognition problem. We show how to find people by finding candidate body segments, and then constructing assemblies of segments that are consistent with the constraints on the appearance of a person that result from kinematic properties. Since a reasonable model of a person requires at least nine segments, it is not possible to inspect every group, due to the huge combinatorial complexity. We propose two approaches to this problem. In one, the search can be pruned by using projected versions of a classifier that accepts groups corresponding to people. We describe an efficient projection algorithm for one popular classifier, and demonstrate that our approach can be used to determine whether images of real scenes contain people. The second approach employs a probabilistic framework, so that we can draw samples of assemblies, with probabilities proportional to their likelihood, which allows to draw human-like assemblies more often than the non-person ones. The main performance problem is in segmentation of images, but the overall results of both approaches on real images of people are encouraging.

AB - Finding people in pictures presents a particularly difficult object recognition problem. We show how to find people by finding candidate body segments, and then constructing assemblies of segments that are consistent with the constraints on the appearance of a person that result from kinematic properties. Since a reasonable model of a person requires at least nine segments, it is not possible to inspect every group, due to the huge combinatorial complexity. We propose two approaches to this problem. In one, the search can be pruned by using projected versions of a classifier that accepts groups corresponding to people. We describe an efficient projection algorithm for one popular classifier, and demonstrate that our approach can be used to determine whether images of real scenes contain people. The second approach employs a probabilistic framework, so that we can draw samples of assemblies, with probabilities proportional to their likelihood, which allows to draw human-like assemblies more often than the non-person ones. The main performance problem is in segmentation of images, but the overall results of both approaches on real images of people are encouraging.

KW - Grouping correspondence search

KW - Human detection

KW - Object recognition

KW - Probabilistic inference

UR - http://www.scopus.com/inward/record.url?scp=0035358875&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035358875&partnerID=8YFLogxK

U2 - 10.1023/A:1011179004708

DO - 10.1023/A:1011179004708

M3 - Article

AN - SCOPUS:0035358875

VL - 43

SP - 45

EP - 68

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

SN - 0920-5691

IS - 1

ER -