Person Re-Identification by Descriptive and Discriminative Classification

Martin Hirzer, Peter Roth, Csaba Beleznai, Horst Bischof

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Abstract

Person re-identification, i.e., recognizing a single person across spatially disjoint cameras, is an important task in visual surveillance. Existing approaches either try to find a suitable description of the appearance or learn a discriminative model. Since these different representational strategies capture a large extent of complementary information we propose to combine both approaches. First, given a specific query, we rank all samples according to a feature-based similarity, where appearance is modeled by a set of region covariance descriptors. Next, a discriminative model is learned using boosting for feature selection, which provides a more specific classifier. The proposed approach is demonstrated on two datasets, where we show that the combination of a generic descriptive statistical model and a discriminatively learned feature-based model attains considerably better results than the individual models alone. In addition, we give a comparison to the state-of-the-art on a publicly available benchmark dataset.
Originalspracheenglisch
TitelProceedings of the Scandinavian Conference on Image Analysis (SCIA)
Herausgeber (Verlag).
Seiten91-102
PublikationsstatusVeröffentlicht - 2011
VeranstaltungScandinavian Conference on Image Analysis - Ystad Saltsjöbad, Schweden
Dauer: 23 Mai 201127 Mai 2011

Konferenz

KonferenzScandinavian Conference on Image Analysis
LandSchweden
OrtYstad Saltsjöbad
Zeitraum23/05/1127/05/11

Fingerprint

Feature extraction
Classifiers
Cameras
Statistical Models

Fields of Expertise

  • Information, Communication & Computing

Dies zitieren

Hirzer, M., Roth, P., Beleznai, C., & Bischof, H. (2011). Person Re-Identification by Descriptive and Discriminative Classification. in Proceedings of the Scandinavian Conference on Image Analysis (SCIA) (S. 91-102). ..

Person Re-Identification by Descriptive and Discriminative Classification. / Hirzer, Martin; Roth, Peter; Beleznai, Csaba; Bischof, Horst.

Proceedings of the Scandinavian Conference on Image Analysis (SCIA). ., 2011. S. 91-102.

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Hirzer, M, Roth, P, Beleznai, C & Bischof, H 2011, Person Re-Identification by Descriptive and Discriminative Classification. in Proceedings of the Scandinavian Conference on Image Analysis (SCIA). ., S. 91-102, Ystad Saltsjöbad, Schweden, 23/05/11.
Hirzer M, Roth P, Beleznai C, Bischof H. Person Re-Identification by Descriptive and Discriminative Classification. in Proceedings of the Scandinavian Conference on Image Analysis (SCIA). . 2011. S. 91-102
Hirzer, Martin ; Roth, Peter ; Beleznai, Csaba ; Bischof, Horst. / Person Re-Identification by Descriptive and Discriminative Classification. Proceedings of the Scandinavian Conference on Image Analysis (SCIA). ., 2011. S. 91-102
@inproceedings{0c72e4043ab14bdd9851ee6d8ff80d6c,
title = "Person Re-Identification by Descriptive and Discriminative Classification",
abstract = "Person re-identification, i.e., recognizing a single person across spatially disjoint cameras, is an important task in visual surveillance. Existing approaches either try to find a suitable description of the appearance or learn a discriminative model. Since these different representational strategies capture a large extent of complementary information we propose to combine both approaches. First, given a specific query, we rank all samples according to a feature-based similarity, where appearance is modeled by a set of region covariance descriptors. Next, a discriminative model is learned using boosting for feature selection, which provides a more specific classifier. The proposed approach is demonstrated on two datasets, where we show that the combination of a generic descriptive statistical model and a discriminatively learned feature-based model attains considerably better results than the individual models alone. In addition, we give a comparison to the state-of-the-art on a publicly available benchmark dataset.",
author = "Martin Hirzer and Peter Roth and Csaba Beleznai and Horst Bischof",
year = "2011",
language = "English",
pages = "91--102",
booktitle = "Proceedings of the Scandinavian Conference on Image Analysis (SCIA)",
publisher = ".",

}

TY - GEN

T1 - Person Re-Identification by Descriptive and Discriminative Classification

AU - Hirzer, Martin

AU - Roth, Peter

AU - Beleznai, Csaba

AU - Bischof, Horst

PY - 2011

Y1 - 2011

N2 - Person re-identification, i.e., recognizing a single person across spatially disjoint cameras, is an important task in visual surveillance. Existing approaches either try to find a suitable description of the appearance or learn a discriminative model. Since these different representational strategies capture a large extent of complementary information we propose to combine both approaches. First, given a specific query, we rank all samples according to a feature-based similarity, where appearance is modeled by a set of region covariance descriptors. Next, a discriminative model is learned using boosting for feature selection, which provides a more specific classifier. The proposed approach is demonstrated on two datasets, where we show that the combination of a generic descriptive statistical model and a discriminatively learned feature-based model attains considerably better results than the individual models alone. In addition, we give a comparison to the state-of-the-art on a publicly available benchmark dataset.

AB - Person re-identification, i.e., recognizing a single person across spatially disjoint cameras, is an important task in visual surveillance. Existing approaches either try to find a suitable description of the appearance or learn a discriminative model. Since these different representational strategies capture a large extent of complementary information we propose to combine both approaches. First, given a specific query, we rank all samples according to a feature-based similarity, where appearance is modeled by a set of region covariance descriptors. Next, a discriminative model is learned using boosting for feature selection, which provides a more specific classifier. The proposed approach is demonstrated on two datasets, where we show that the combination of a generic descriptive statistical model and a discriminatively learned feature-based model attains considerably better results than the individual models alone. In addition, we give a comparison to the state-of-the-art on a publicly available benchmark dataset.

M3 - Conference contribution

SP - 91

EP - 102

BT - Proceedings of the Scandinavian Conference on Image Analysis (SCIA)

PB - .

ER -