Robust visual pedestrian detection by tight coupling to tracking

Abstract : In this article, we propose a visual pedestrian detection system which couples pedestrian appearance and pedestrian motion in a Bayesian fashion, with the goal of making detection more invariant to appearance changes. More precisely, the system couples dense appearance-based pedes-trian likelihoods derived from a sliding-window SVM detector to spatial prior distributions obtained from the prediction step of a particle filter based pedestrian tracker. This mechanism, which we term dynamic attention priors (DAP), is inspired by recent results on predictive visual attention in humans and can be implemented at negligible computational cost. We prove experimentally, using a set of public, annotated pedestrian sequences, that detection performance is improved significantly, especially in cases where pedestrians differ from the learned models, e.g., when they are too small, have an unusual pose or occur before strongly structured backgrounds. In particular, dynamic attention priors allow to use more restrictive detection thresholds without losing detections while minimizing false detections.
Type de document :
Communication dans un congrès
IEEE International Conference On Intelligent Transportation Systems (ITSC), Oct 2014, Qingdao, China. pp.1935 - 1940, 2014, 〈10.1109/ITSC.2014.6957989〉
Liste complète des métadonnées

Littérature citée [30 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01098703
Contributeur : Alexander Gepperth <>
Soumis le : dimanche 28 décembre 2014 - 15:57:42
Dernière modification le : mercredi 28 mars 2018 - 13:25:26
Document(s) archivé(s) le : mercredi 3 juin 2015 - 12:35:25

Fichier

gepperth2014robust.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Alexander Gepperth, Egor Sattarov, Bernd Heisele, Sergio Alberto Rodriguez Florez. Robust visual pedestrian detection by tight coupling to tracking. IEEE International Conference On Intelligent Transportation Systems (ITSC), Oct 2014, Qingdao, China. pp.1935 - 1940, 2014, 〈10.1109/ITSC.2014.6957989〉. 〈hal-01098703〉

Partager

Métriques

Consultations de la notice

355

Téléchargements de fichiers

498