Applications of pathfinder network scaling for improving the ranking of satellite images

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Content-based image retrieval techniques, although promising for handling large quantities of geospatial image data, are prone to creating overfitted models. This is due to the fact that supervised models most often capture patterns of existing observations and not those related to the whole population. This results in models that do not generalize well to new, undiscovered images. This article proposes a methodology to reduce overfitting when ranking high-resolution satellite images by domain semantics. Our approach uses PathFinder Network Scaling ensemble methods. We generate cross-fold co-occurrence measures for relevance of feature subspaces to each semantic. Each matrix is then reduced using the PathFinder network scaling algorithm. Irrelevant nodes are removed using node strength metrics resulting in an optimized model for ranking by semantic that generalizes better to new images. The results show that, when using this approach, the quality of ranking by semantic can be significantly improved. Mean Average Precision (MAP) of ranking over cross-fold experiments increased by 13.2% while standard deviation of MAP was reduced by 16.8% relative to experiments without PathFinder network scaling.

Original languageEnglish (US)
Article number6461103
Pages (from-to)1092-1099
Number of pages8
JournalIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume6
Issue number3
DOIs
StatePublished - Feb 15 2013

Fingerprint

ranking
Semantics
Satellites
fold
Image retrieval
experiment
Experiments
matrix
methodology
satellite image

All Science Journal Classification (ASJC) codes

  • Computers in Earth Sciences
  • Atmospheric Science

Cite this

@article{d4b7b8f1f8504909b9243b7cc0e6ec8f,
title = "Applications of pathfinder network scaling for improving the ranking of satellite images",
abstract = "Content-based image retrieval techniques, although promising for handling large quantities of geospatial image data, are prone to creating overfitted models. This is due to the fact that supervised models most often capture patterns of existing observations and not those related to the whole population. This results in models that do not generalize well to new, undiscovered images. This article proposes a methodology to reduce overfitting when ranking high-resolution satellite images by domain semantics. Our approach uses PathFinder Network Scaling ensemble methods. We generate cross-fold co-occurrence measures for relevance of feature subspaces to each semantic. Each matrix is then reduced using the PathFinder network scaling algorithm. Irrelevant nodes are removed using node strength metrics resulting in an optimized model for ranking by semantic that generalizes better to new images. The results show that, when using this approach, the quality of ranking by semantic can be significantly improved. Mean Average Precision (MAP) of ranking over cross-fold experiments increased by 13.2{\%} while standard deviation of MAP was reduced by 16.8{\%} relative to experiments without PathFinder network scaling.",
author = "Barb, {Adrian Sorin} and Roy Clariana and Shyu, {Chi Ren}",
year = "2013",
month = "2",
day = "15",
doi = "10.1109/JSTARS.2013.2242254",
language = "English (US)",
volume = "6",
pages = "1092--1099",
journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",
issn = "1939-1404",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Applications of pathfinder network scaling for improving the ranking of satellite images

AU - Barb, Adrian Sorin

AU - Clariana, Roy

AU - Shyu, Chi Ren

PY - 2013/2/15

Y1 - 2013/2/15

N2 - Content-based image retrieval techniques, although promising for handling large quantities of geospatial image data, are prone to creating overfitted models. This is due to the fact that supervised models most often capture patterns of existing observations and not those related to the whole population. This results in models that do not generalize well to new, undiscovered images. This article proposes a methodology to reduce overfitting when ranking high-resolution satellite images by domain semantics. Our approach uses PathFinder Network Scaling ensemble methods. We generate cross-fold co-occurrence measures for relevance of feature subspaces to each semantic. Each matrix is then reduced using the PathFinder network scaling algorithm. Irrelevant nodes are removed using node strength metrics resulting in an optimized model for ranking by semantic that generalizes better to new images. The results show that, when using this approach, the quality of ranking by semantic can be significantly improved. Mean Average Precision (MAP) of ranking over cross-fold experiments increased by 13.2% while standard deviation of MAP was reduced by 16.8% relative to experiments without PathFinder network scaling.

AB - Content-based image retrieval techniques, although promising for handling large quantities of geospatial image data, are prone to creating overfitted models. This is due to the fact that supervised models most often capture patterns of existing observations and not those related to the whole population. This results in models that do not generalize well to new, undiscovered images. This article proposes a methodology to reduce overfitting when ranking high-resolution satellite images by domain semantics. Our approach uses PathFinder Network Scaling ensemble methods. We generate cross-fold co-occurrence measures for relevance of feature subspaces to each semantic. Each matrix is then reduced using the PathFinder network scaling algorithm. Irrelevant nodes are removed using node strength metrics resulting in an optimized model for ranking by semantic that generalizes better to new images. The results show that, when using this approach, the quality of ranking by semantic can be significantly improved. Mean Average Precision (MAP) of ranking over cross-fold experiments increased by 13.2% while standard deviation of MAP was reduced by 16.8% relative to experiments without PathFinder network scaling.

UR - http://www.scopus.com/inward/record.url?scp=84880291114&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84880291114&partnerID=8YFLogxK

U2 - 10.1109/JSTARS.2013.2242254

DO - 10.1109/JSTARS.2013.2242254

M3 - Article

VL - 6

SP - 1092

EP - 1099

JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

SN - 1939-1404

IS - 3

M1 - 6461103

ER -