Discovering Primitive Action Categories by Leveraging Relevant Visual Context

Abstract : Under the bag-of-features framework we aim to learn primitive action categories from video without supervision by leveraging relevant visual context in addition to motion features. We define visual context as the appearance of the entire scene including the actor, related objects and relevant background features. To leverage visual context along with motion features, we learn a bi-modal latent variable model to discover action categories without supervision. Our experiments show that the combination of relevant visual context and motion features improves the performance of action discovery. Furthermore, we show that our method is able to leverage relevant visual features for action discovery despite the presence of irrelevant background objects.
Type de document :
Communication dans un congrès
The Eighth International Workshop on Visual Surveillance - VS2008, Oct 2008, Marseille, France. 2008
Liste complète des métadonnées

Littérature citée [30 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/inria-00325777
Contributeur : Peter Sturm <>
Soumis le : mardi 30 septembre 2008 - 11:28:59
Dernière modification le : samedi 16 décembre 2017 - 06:54:03
Document(s) archivé(s) le : vendredi 4 juin 2010 - 12:00:04

Fichier

VS2008-Poster-a.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : inria-00325777, version 1

Collections

Citation

Kris M. Kitani, Takahiro Okabe, Yoichi Sato, Akihiro Sugimoto. Discovering Primitive Action Categories by Leveraging Relevant Visual Context. The Eighth International Workshop on Visual Surveillance - VS2008, Oct 2008, Marseille, France. 2008. 〈inria-00325777〉

Partager

Métriques

Consultations de la notice

129

Téléchargements de fichiers

144