Hierarchical Two-pathway Autoencoders Neural Networks for Skyline Context Conceptualization - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue International Journal of Information and Decision Sciences Année : 2020

Hierarchical Two-pathway Autoencoders Neural Networks for Skyline Context Conceptualization

Ameni Sassi
  • Fonction : Auteur
  • PersonId : 1044258
Wael Ouarda
  • Fonction : Auteur
  • PersonId : 1044259
Chokri Ben Amar
  • Fonction : Auteur
  • PersonId : 991123

Résumé

In this paper, we proposed a novel hierarchical two-pathway autoencoders architecture to transform a local information based on skyline scene representation, into non-linear space. The first pathway is intended for the transformation of the geometric features extracted from the horizon line. The second pathway is applied after the first one to joint the color information under the skyline to the transformed geometric features, and to get the landscape context conceptualization. To evaluate our suggested system, we constructed the SKYLINEScene database containing 2000 images of rural and urban landscapes, with a high degree of diversity. In order to investigate the performance of our HTANN-Skyline, many experiments were carried out using this new database. Our approach shows its robustness in Skyline context conceptualization and enhances the classification rates by 1% compared to the AlexNet architecture; and by more than 10% compared to the hand-crafted approaches based on global and local features. at the Université Lumière Lyon 2, and a member of the LIRIS laboratory, UMR CNRS 5205. His main research activities are devoted to models and tools for image processing, image analysis, shape recognition.
Fichier non déposé

Dates et versions

hal-02072150 , version 1 (13-06-2019)

Identifiants

Citer

Ameni Sassi, Wael Ouarda, Chokri Ben Amar, Serge Miguet. Hierarchical Two-pathway Autoencoders Neural Networks for Skyline Context Conceptualization. International Journal of Information and Decision Sciences, 2020, 12 (4), pp.299 - 327. ⟨10.1504/IJIDS.2020.110447⟩. ⟨hal-02072150⟩
215 Consultations
1 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More