Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos.

Barbara André, Tom Vercauteren, Anna M. Buchner, Michael B Wallace, Nicholas Ayache

Research output: Chapter in Book/Report/Conference proceedingChapter

9 Citations (Scopus)

Abstract

Evaluating content-based retrieval (CBR) is challenging because it requires an adequate ground-truth. When the available groundtruth is limited to textual metadata such as pathological classes, retrieval results can only be evaluated indirectly, for example in terms of classification performance. In this study we first present a tool to generate perceived similarity ground-truth that enables direct evaluation of endomicroscopic video retrieval. This tool uses a four-points Likert scale and collects subjective pairwise similarities perceived by multiple expert observers. We then evaluate against the generated ground-truth a previously developed dense bag-of-visual-words method for endomicroscopic video retrieval. Confirming the results of previous indirect evaluation based on classification, our direct evaluation shows that this method significantly outperforms several other state-of-the-art CBR methods. In a second step, we propose to improve the CBR method by learning an adjusted similarity metric from the perceived similarity ground-truth. By minimizing a margin-based cost function that differentiates similar and dissimilar video pairs, we learn a weight vector applied to the visual word signatures of videos. Using cross-validation, we demonstrate that the learned similarity distance is significantly better correlated with the perceived similarity than the original visual-word-based distance.

Original languageEnglish (US)
Title of host publicationMedical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
Pages297-304
Number of pages8
Volume14
EditionPt 3
StatePublished - 2011

Fingerprint

Distance Education
Learning
Weights and Measures
Costs and Cost Analysis

ASJC Scopus subject areas

  • Medicine(all)

Cite this

André, B., Vercauteren, T., Buchner, A. M., Wallace, M. B., & Ayache, N. (2011). Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos. In Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention (Pt 3 ed., Vol. 14, pp. 297-304)

Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos. / André, Barbara; Vercauteren, Tom; Buchner, Anna M.; Wallace, Michael B; Ayache, Nicholas.

Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Vol. 14 Pt 3. ed. 2011. p. 297-304.

Research output: Chapter in Book/Report/Conference proceedingChapter

André, B, Vercauteren, T, Buchner, AM, Wallace, MB & Ayache, N 2011, Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos. in Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Pt 3 edn, vol. 14, pp. 297-304.
André B, Vercauteren T, Buchner AM, Wallace MB, Ayache N. Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos. In Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Pt 3 ed. Vol. 14. 2011. p. 297-304
André, Barbara ; Vercauteren, Tom ; Buchner, Anna M. ; Wallace, Michael B ; Ayache, Nicholas. / Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos. Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Vol. 14 Pt 3. ed. 2011. pp. 297-304
@inbook{4ea7bde281a24bd297d14dd23375e204,
title = "Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos.",
abstract = "Evaluating content-based retrieval (CBR) is challenging because it requires an adequate ground-truth. When the available groundtruth is limited to textual metadata such as pathological classes, retrieval results can only be evaluated indirectly, for example in terms of classification performance. In this study we first present a tool to generate perceived similarity ground-truth that enables direct evaluation of endomicroscopic video retrieval. This tool uses a four-points Likert scale and collects subjective pairwise similarities perceived by multiple expert observers. We then evaluate against the generated ground-truth a previously developed dense bag-of-visual-words method for endomicroscopic video retrieval. Confirming the results of previous indirect evaluation based on classification, our direct evaluation shows that this method significantly outperforms several other state-of-the-art CBR methods. In a second step, we propose to improve the CBR method by learning an adjusted similarity metric from the perceived similarity ground-truth. By minimizing a margin-based cost function that differentiates similar and dissimilar video pairs, we learn a weight vector applied to the visual word signatures of videos. Using cross-validation, we demonstrate that the learned similarity distance is significantly better correlated with the perceived similarity than the original visual-word-based distance.",
author = "Barbara Andr{\'e} and Tom Vercauteren and Buchner, {Anna M.} and Wallace, {Michael B} and Nicholas Ayache",
year = "2011",
language = "English (US)",
volume = "14",
pages = "297--304",
booktitle = "Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention",
edition = "Pt 3",

}

TY - CHAP

T1 - Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos.

AU - André, Barbara

AU - Vercauteren, Tom

AU - Buchner, Anna M.

AU - Wallace, Michael B

AU - Ayache, Nicholas

PY - 2011

Y1 - 2011

N2 - Evaluating content-based retrieval (CBR) is challenging because it requires an adequate ground-truth. When the available groundtruth is limited to textual metadata such as pathological classes, retrieval results can only be evaluated indirectly, for example in terms of classification performance. In this study we first present a tool to generate perceived similarity ground-truth that enables direct evaluation of endomicroscopic video retrieval. This tool uses a four-points Likert scale and collects subjective pairwise similarities perceived by multiple expert observers. We then evaluate against the generated ground-truth a previously developed dense bag-of-visual-words method for endomicroscopic video retrieval. Confirming the results of previous indirect evaluation based on classification, our direct evaluation shows that this method significantly outperforms several other state-of-the-art CBR methods. In a second step, we propose to improve the CBR method by learning an adjusted similarity metric from the perceived similarity ground-truth. By minimizing a margin-based cost function that differentiates similar and dissimilar video pairs, we learn a weight vector applied to the visual word signatures of videos. Using cross-validation, we demonstrate that the learned similarity distance is significantly better correlated with the perceived similarity than the original visual-word-based distance.

AB - Evaluating content-based retrieval (CBR) is challenging because it requires an adequate ground-truth. When the available groundtruth is limited to textual metadata such as pathological classes, retrieval results can only be evaluated indirectly, for example in terms of classification performance. In this study we first present a tool to generate perceived similarity ground-truth that enables direct evaluation of endomicroscopic video retrieval. This tool uses a four-points Likert scale and collects subjective pairwise similarities perceived by multiple expert observers. We then evaluate against the generated ground-truth a previously developed dense bag-of-visual-words method for endomicroscopic video retrieval. Confirming the results of previous indirect evaluation based on classification, our direct evaluation shows that this method significantly outperforms several other state-of-the-art CBR methods. In a second step, we propose to improve the CBR method by learning an adjusted similarity metric from the perceived similarity ground-truth. By minimizing a margin-based cost function that differentiates similar and dissimilar video pairs, we learn a weight vector applied to the visual word signatures of videos. Using cross-validation, we demonstrate that the learned similarity distance is significantly better correlated with the perceived similarity than the original visual-word-based distance.

UR - http://www.scopus.com/inward/record.url?scp=82255164550&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=82255164550&partnerID=8YFLogxK

M3 - Chapter

C2 - 22003712

AN - SCOPUS:82255164550

VL - 14

SP - 297

EP - 304

BT - Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention

ER -