An ensemble color model for human re-identification

Xiaokai Liu, Hongyu Wang, Yi Wu, Jimei Yang, Ming Hsuan Yang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

39 Citations (Scopus)

Abstract

Appearance-based human re-identification is challenging due to different camera characteristics, varying lighting conditions, pose variations across camera views, etc. Recent studies have revealed that color information plays a critical role on performance. However, two problems remain unclear: (1) how do different color descriptors perform under the same scene in re-identification problem? and (2) how can we combine these descriptors without losing their invariance property and distinctiveness power? In this paper, we propose a novel ensemble model that combines different color descriptors in the decision level through metric learning. Experiments show that the proposed system significantly outperforms state-of-the-art algorithms on two challenging datasets (VIPeR and PRID 450S). We have improved the Rank 1 recognition rate on VIPeR dataset by 8.7%.

Original languageEnglish
Title of host publicationProceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages868-875
Number of pages8
ISBN (Electronic)9781479966820
DOIs
Publication statusPublished - 2015 Feb 19
Event2015 15th IEEE Winter Conference on Applications of Computer Vision, WACV 2015 - Waikoloa, United States
Duration: 2015 Jan 52015 Jan 9

Publication series

NameProceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015

Conference

Conference2015 15th IEEE Winter Conference on Applications of Computer Vision, WACV 2015
CountryUnited States
CityWaikoloa
Period15/1/515/1/9

Fingerprint

Identification (control systems)
Color
Cameras
Invariance
Lighting
Experiments

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Computer Vision and Pattern Recognition

Cite this

Liu, X., Wang, H., Wu, Y., Yang, J., & Yang, M. H. (2015). An ensemble color model for human re-identification. In Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015 (pp. 868-875). [7045974] (Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WACV.2015.120
Liu, Xiaokai ; Wang, Hongyu ; Wu, Yi ; Yang, Jimei ; Yang, Ming Hsuan. / An ensemble color model for human re-identification. Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015. Institute of Electrical and Electronics Engineers Inc., 2015. pp. 868-875 (Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015).
@inproceedings{e20a5db3345f4457a60fe8a833709ac5,
title = "An ensemble color model for human re-identification",
abstract = "Appearance-based human re-identification is challenging due to different camera characteristics, varying lighting conditions, pose variations across camera views, etc. Recent studies have revealed that color information plays a critical role on performance. However, two problems remain unclear: (1) how do different color descriptors perform under the same scene in re-identification problem? and (2) how can we combine these descriptors without losing their invariance property and distinctiveness power? In this paper, we propose a novel ensemble model that combines different color descriptors in the decision level through metric learning. Experiments show that the proposed system significantly outperforms state-of-the-art algorithms on two challenging datasets (VIPeR and PRID 450S). We have improved the Rank 1 recognition rate on VIPeR dataset by 8.7{\%}.",
author = "Xiaokai Liu and Hongyu Wang and Yi Wu and Jimei Yang and Yang, {Ming Hsuan}",
year = "2015",
month = "2",
day = "19",
doi = "10.1109/WACV.2015.120",
language = "English",
series = "Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "868--875",
booktitle = "Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015",
address = "United States",

}

Liu, X, Wang, H, Wu, Y, Yang, J & Yang, MH 2015, An ensemble color model for human re-identification. in Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015., 7045974, Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015, Institute of Electrical and Electronics Engineers Inc., pp. 868-875, 2015 15th IEEE Winter Conference on Applications of Computer Vision, WACV 2015, Waikoloa, United States, 15/1/5. https://doi.org/10.1109/WACV.2015.120

An ensemble color model for human re-identification. / Liu, Xiaokai; Wang, Hongyu; Wu, Yi; Yang, Jimei; Yang, Ming Hsuan.

Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015. Institute of Electrical and Electronics Engineers Inc., 2015. p. 868-875 7045974 (Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - An ensemble color model for human re-identification

AU - Liu, Xiaokai

AU - Wang, Hongyu

AU - Wu, Yi

AU - Yang, Jimei

AU - Yang, Ming Hsuan

PY - 2015/2/19

Y1 - 2015/2/19

N2 - Appearance-based human re-identification is challenging due to different camera characteristics, varying lighting conditions, pose variations across camera views, etc. Recent studies have revealed that color information plays a critical role on performance. However, two problems remain unclear: (1) how do different color descriptors perform under the same scene in re-identification problem? and (2) how can we combine these descriptors without losing their invariance property and distinctiveness power? In this paper, we propose a novel ensemble model that combines different color descriptors in the decision level through metric learning. Experiments show that the proposed system significantly outperforms state-of-the-art algorithms on two challenging datasets (VIPeR and PRID 450S). We have improved the Rank 1 recognition rate on VIPeR dataset by 8.7%.

AB - Appearance-based human re-identification is challenging due to different camera characteristics, varying lighting conditions, pose variations across camera views, etc. Recent studies have revealed that color information plays a critical role on performance. However, two problems remain unclear: (1) how do different color descriptors perform under the same scene in re-identification problem? and (2) how can we combine these descriptors without losing their invariance property and distinctiveness power? In this paper, we propose a novel ensemble model that combines different color descriptors in the decision level through metric learning. Experiments show that the proposed system significantly outperforms state-of-the-art algorithms on two challenging datasets (VIPeR and PRID 450S). We have improved the Rank 1 recognition rate on VIPeR dataset by 8.7%.

UR - http://www.scopus.com/inward/record.url?scp=84925385115&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84925385115&partnerID=8YFLogxK

U2 - 10.1109/WACV.2015.120

DO - 10.1109/WACV.2015.120

M3 - Conference contribution

AN - SCOPUS:84925385115

T3 - Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015

SP - 868

EP - 875

BT - Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Liu X, Wang H, Wu Y, Yang J, Yang MH. An ensemble color model for human re-identification. In Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015. Institute of Electrical and Electronics Engineers Inc. 2015. p. 868-875. 7045974. (Proceedings - 2015 IEEE Winter Conference on Applications of Computer Vision, WACV 2015). https://doi.org/10.1109/WACV.2015.120