Video retrieval with CNN features - Archive ouverte HAL
Communication Dans Un Congrès Année : 2019

Video retrieval with CNN features

Résumé

Convolutional neural network features are becoming the norm in instance retrieval. This work investigate the relevance of using an of the shelf object detection network like Faster R-CNN as a feature extractor. We build an Image-to-video face retrieval pipeline composed of filtering and re-ranking that uses the objects proposals learned by a Region Proposal Network (RPN) and their associated representations taken from a CNN. Moreover we study the relevance of features from a finetuned network. The results obtained are very promising
Fichier principal
Vignette du fichier
COC.pdf (182.72 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02296746 , version 1 (25-09-2019)

Identifiants

  • HAL Id : hal-02296746 , version 1

Citer

Imane Hachchane, Abdelmajid Badri, Aïcha Sahel, Yassine Ruichek. Video retrieval with CNN features. Colloque sur les Objets et systèmes Connectés, Ecole Supérieure de Technologie de Casablanca (Maroc), Institut Universitaire de Technologie d'Aix-Marseille (France), Jun 2019, CASABLANCA, Morocco. ⟨hal-02296746⟩
142 Consultations
102 Téléchargements

Partager

More