Error metric analysis and its applications

Qi Tian, Qing Xue, Nicu Sebe, Thomas S Huang

Research output: Contribution to journalConference article

Abstract

In this paper, we present a general guideline to establish the relation of noise distribution model and its corresponding error metric. By designing error metrics, we obtain a much richer set of distance measures besides the conventional Euclidean distance or SSD (sum of the squared difference) and the Manhattan distance or SAD (sum of the absolute difference). The corresponding nonlinear estimations such as harmonic mean, geometric mean, as well as their generalized nonlinear operations are derived. It not only offers more flexibility than the conventional metrics but also discloses the coherent relation between the noise model and its corresponding error metric. We experiment with different error metrics for similarity noise estimation and compute the accuracy of different methods in three kinds of applications: content-based image retrieval from a large database, stereo matching, and motion tracking in video sequences. In all the experiments, robust results are obtained for noise estimation based on the proposed error metric analysis.

Original languageEnglish (US)
Article number05
Pages (from-to)46-57
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume5601
DOIs
StatePublished - Dec 1 2004
EventInternet Multimedia Management Systems V - Philadelphia, PA, United States
Duration: Oct 26 2004Oct 28 2004

Fingerprint

Metric
Noise Estimation
Image retrieval
Harmonic mean
Nonlinear Estimation
Motion Tracking
Stereo Matching
Content-based Image Retrieval
Geometric mean
retrieval
Distance Measure
Euclidean Distance
flexibility
Experiments
Experiment
harmonics
Flexibility
Model

Keywords

  • Content-based image retrieval
  • Maximum likelihood
  • SAD
  • SSD
  • Similarity noise distribution
  • Stereo matching

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Error metric analysis and its applications. / Tian, Qi; Xue, Qing; Sebe, Nicu; Huang, Thomas S.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 5601, 05, 01.12.2004, p. 46-57.

Research output: Contribution to journalConference article

@article{02cb34edf41b4b60be214ab49cb478a4,
title = "Error metric analysis and its applications",
abstract = "In this paper, we present a general guideline to establish the relation of noise distribution model and its corresponding error metric. By designing error metrics, we obtain a much richer set of distance measures besides the conventional Euclidean distance or SSD (sum of the squared difference) and the Manhattan distance or SAD (sum of the absolute difference). The corresponding nonlinear estimations such as harmonic mean, geometric mean, as well as their generalized nonlinear operations are derived. It not only offers more flexibility than the conventional metrics but also discloses the coherent relation between the noise model and its corresponding error metric. We experiment with different error metrics for similarity noise estimation and compute the accuracy of different methods in three kinds of applications: content-based image retrieval from a large database, stereo matching, and motion tracking in video sequences. In all the experiments, robust results are obtained for noise estimation based on the proposed error metric analysis.",
keywords = "Content-based image retrieval, Maximum likelihood, SAD, SSD, Similarity noise distribution, Stereo matching",
author = "Qi Tian and Qing Xue and Nicu Sebe and Huang, {Thomas S}",
year = "2004",
month = "12",
day = "1",
doi = "10.1117/12.571723",
language = "English (US)",
volume = "5601",
pages = "46--57",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

TY - JOUR

T1 - Error metric analysis and its applications

AU - Tian, Qi

AU - Xue, Qing

AU - Sebe, Nicu

AU - Huang, Thomas S

PY - 2004/12/1

Y1 - 2004/12/1

N2 - In this paper, we present a general guideline to establish the relation of noise distribution model and its corresponding error metric. By designing error metrics, we obtain a much richer set of distance measures besides the conventional Euclidean distance or SSD (sum of the squared difference) and the Manhattan distance or SAD (sum of the absolute difference). The corresponding nonlinear estimations such as harmonic mean, geometric mean, as well as their generalized nonlinear operations are derived. It not only offers more flexibility than the conventional metrics but also discloses the coherent relation between the noise model and its corresponding error metric. We experiment with different error metrics for similarity noise estimation and compute the accuracy of different methods in three kinds of applications: content-based image retrieval from a large database, stereo matching, and motion tracking in video sequences. In all the experiments, robust results are obtained for noise estimation based on the proposed error metric analysis.

AB - In this paper, we present a general guideline to establish the relation of noise distribution model and its corresponding error metric. By designing error metrics, we obtain a much richer set of distance measures besides the conventional Euclidean distance or SSD (sum of the squared difference) and the Manhattan distance or SAD (sum of the absolute difference). The corresponding nonlinear estimations such as harmonic mean, geometric mean, as well as their generalized nonlinear operations are derived. It not only offers more flexibility than the conventional metrics but also discloses the coherent relation between the noise model and its corresponding error metric. We experiment with different error metrics for similarity noise estimation and compute the accuracy of different methods in three kinds of applications: content-based image retrieval from a large database, stereo matching, and motion tracking in video sequences. In all the experiments, robust results are obtained for noise estimation based on the proposed error metric analysis.

KW - Content-based image retrieval

KW - Maximum likelihood

KW - SAD

KW - SSD

KW - Similarity noise distribution

KW - Stereo matching

UR - http://www.scopus.com/inward/record.url?scp=17644420350&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=17644420350&partnerID=8YFLogxK

U2 - 10.1117/12.571723

DO - 10.1117/12.571723

M3 - Conference article

AN - SCOPUS:17644420350

VL - 5601

SP - 46

EP - 57

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

M1 - 05

ER -