Human Re-identification System On Highly Parallel GPU and CPU Architectures

Abstract : The paper presents a new approach to the human reindetification problem using covariance features. In many cases, a distance operator between signatures, based on generalized eigenvalues, has to be computed efficiently, especially once the real-time response time is expected from the system. This is a challenging problem as many procedures are in fact computationally intensive tasks and must be repeated constantly. To deal with this problem we have successfully designed and then tested a new video surveillance system. To obtain the required high efficiency we took the advantage of highly parallel computing architectures such as FPGA, GPU and CPU units to perform calculations. However, we had to propose a new GPU-based implementation of the distance operator for querying the example database. Thus, in this paper we present experimental evaluation of the proposed solution in the light of the database response time depending on its size.
Type de document :
Communication dans un congrès
Dziech, Andrzej and Czyżewski, Andrzej. Multimedia Communications, Services and Security, Jun 2011, Krakow, Poland. Springer Berlin Heidelberg, 149, 2011, Communications in Computer and Information Science. 〈10.1007/978-3-642-21512-4_35〉
Liste complète des métadonnées

Littérature citée [10 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-00645938
Contributeur : Slawomir Bak <>
Soumis le : mardi 6 décembre 2011 - 11:02:22
Dernière modification le : jeudi 11 janvier 2018 - 17:07:46
Document(s) archivé(s) le : vendredi 16 novembre 2012 - 14:25:57

Fichier

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

Identifiants

Collections

Citation

Slawomir Bak, Krzysztof Kurowski, Krystyna Napierala. Human Re-identification System On Highly Parallel GPU and CPU Architectures. Dziech, Andrzej and Czyżewski, Andrzej. Multimedia Communications, Services and Security, Jun 2011, Krakow, Poland. Springer Berlin Heidelberg, 149, 2011, Communications in Computer and Information Science. 〈10.1007/978-3-642-21512-4_35〉. 〈hal-00645938〉

Partager

Métriques

Consultations de la notice

270

Téléchargements de fichiers

200