Information Theoretic Rotationwise Robust Binary Descriptor Learning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Information Theoretic Rotationwise Robust Binary Descriptor Learning

Un descripteur binaire robuste aux rotations basé sur la théorie de l'information

Résumé

In this paper, we propose a new data-driven approach for binary descriptor selection. In order to draw a clear analysis of common designs, we present a general information-theoretic selection paradigm. It encompasses several standard binary descriptor construction schemes, including a recent state-of-the-art one named BOLD. We pursue the same endeavor to increase the stability of the produced descriptors with respect to rotations. To achieve this goal, we have designed a novel of-fline selection criterion which is better adapted to the online matching procedure. The effectiveness of our approach is demonstrated on two standard datasets, where our descriptor is compared to BOLD and to several classical descriptors. In particular, it emerges that our approach can reproduce equivalent if not better performance as BOLD while relying on twice shorter descriptors. Such an improvement can be influential for real-time applications.

Mots clés

Fichier principal
Vignette du fichier
youssef.pdf (484.4 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-01418934 , version 1 (17-12-2016)

Identifiants

  • HAL Id : hal-01418934 , version 1

Citer

Youssef El Rhabi, Loïc Simon, Luc Brun, Josep Llados, Felipe Lumbreras. Information Theoretic Rotationwise Robust Binary Descriptor Learning. Structural, Syntactic, and Statistical Pattern Recognition, Nov 2016, Mérida, Mexico. pp.368--378. ⟨hal-01418934⟩
105 Consultations
176 Téléchargements

Partager

Gmail Facebook X LinkedIn More