Multiple Object Tracking by Efficient Graph Partitioning

Ratnesh Kumar 1 Guillaume Charpiat 1 Monique Thonnat 1
1 STARS - Spatio-Temporal Activity Recognition Systems
CRISAM - Inria Sophia Antipolis - Méditerranée
Abstract : In this paper, we view multiple object tracking as a graphpartitioning problem. Given any object detector, we build the graph ofall detections and aim to partition it into trajectories. To quantifythe similarity of any two detections, we consider local cues such as pointtracks and speed, global cues such as appearance, as well as intermediateones such as trajectory straightness. These different clues are dealt jointlyto make the approach robust to detection mistakes (missing or extradetections). We thus define a Conditional Random Field and optimizeit using an efficient combination of message passing and move-makingalgorithms. Our approach is fast on video batch sizes of hundreds offrames. Competitive and stable results on varied videos demonstrate the robustnessand efficiency of our approach.
Type de document :
Communication dans un congrès
Terence Sim and Jianxin Wu. ACCV - 12th Asian Conference on Computer Vision, Nov 2014, Singapore, Singapore. Springer, 2014, Lecture Notes on Computer Science
Liste complète des métadonnées

Littérature citée [26 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01061450
Contributeur : Guillaume Charpiat <>
Soumis le : mercredi 19 novembre 2014 - 17:13:23
Dernière modification le : samedi 2 décembre 2017 - 01:27:36
Document(s) archivé(s) le : vendredi 20 février 2015 - 11:10:45

Fichier

ACCV2014_Tracking.pdf
Fichiers produits par l'(les) auteur(s)

Licence


Distributed under a Creative Commons Paternité - Pas d'utilisation commerciale 4.0 International License

Identifiants

  • HAL Id : hal-01061450, version 1

Collections

Citation

Ratnesh Kumar, Guillaume Charpiat, Monique Thonnat. Multiple Object Tracking by Efficient Graph Partitioning. Terence Sim and Jianxin Wu. ACCV - 12th Asian Conference on Computer Vision, Nov 2014, Singapore, Singapore. Springer, 2014, Lecture Notes on Computer Science. 〈hal-01061450〉

Partager

Métriques

Consultations de la notice

312

Téléchargements de fichiers

204