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 ground-truth 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.
Type de document :
Communication dans un congrès
Medical Image Computing and Computer-Assisted Intervention - MICCAI 2011, 2011, Toronto, Canada. 14, pp.297-304, 2011, Pt 3. 〈10.1007/978-3-642-23626-6_37〉
Liste complète des métadonnées

Littérature citée [11 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/inria-00598301
Contributeur : Barbara André <>
Soumis le : lundi 6 juin 2011 - 10:26:18
Dernière modification le : jeudi 11 janvier 2018 - 16:19:58
Document(s) archivé(s) le : mercredi 7 septembre 2011 - 09:51:59

Fichiers

Barbara_ANDRE_MICCAI11_CameraR...
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Barbara André, Tom Vercauteren, Anna Buchner, Michael Wallace, Nicholas Ayache. Retrieval evaluation and distance learning from perceived similarity between endomicroscopy videos. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2011, 2011, Toronto, Canada. 14, pp.297-304, 2011, Pt 3. 〈10.1007/978-3-642-23626-6_37〉. 〈inria-00598301〉

Partager

Métriques

Consultations de la notice

271

Téléchargements de fichiers

162