The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
La plupart des approches de suivi discriminant de pointe sont basées soit sur des modèles d'apparence de modèles, soit sur des modèles d'apparence statistiques. Bien que les modèles d’apparence de modèle aient montré d’excellentes performances, ils fonctionnent mal lorsque l’apparence de la cible change rapidement. En revanche, les modèles d'apparence statistique sont insensibles aux changements rapides d'état de la cible, mais ils donnent des résultats de suivi inférieurs dans des scénarios difficiles tels que les variations d'éclairage et les parasites d'arrière-plan. Dans cet article, nous proposons une approche adaptative de suivi d'objets avec des modèles complémentaires basés sur des modèles de modèles et d'apparence statistique. Ces deux modèles sont unifiés via notre nouvelle stratégie de combinaison. De plus, nous introduisons un schéma de mise à jour efficace pour améliorer les performances de notre approche. Les résultats expérimentaux démontrent que notre approche permet d'obtenir des performances supérieures à des vitesses qui dépassent de loin les exigences de fréquence d'images sur les récents tests de suivi.
Peng GAO
Harbin Institute of Technology
Yipeng MA
Harbin Institute of Technology
Chao LI
Harbin Institute of Technology
Ke SONG
Harbin Institute of Technology
Yan ZHANG
Harbin Institute of Technology
Fei WANG
Harbin Institute of Technology
Liyi XIAO
Harbin Institute of Technology
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Peng GAO, Yipeng MA, Chao LI, Ke SONG, Yan ZHANG, Fei WANG, Liyi XIAO, "Adaptive Object Tracking with Complementary Models" in IEICE TRANSACTIONS on Information,
vol. E101-D, no. 11, pp. 2849-2854, November 2018, doi: 10.1587/transinf.2018EDL8074.
Abstract: Most state-of-the-art discriminative tracking approaches are based on either template appearance models or statistical appearance models. Despite template appearance models have shown excellent performance, they perform poorly when the target appearance changes rapidly. In contrast, statistic appearance models are insensitive to fast target state changes, but they yield inferior tracking results in challenging scenarios such as illumination variations and background clutters. In this paper, we propose an adaptive object tracking approach with complementary models based on template and statistical appearance models. Both of these models are unified via our novel combination strategy. In addition, we introduce an efficient update scheme to improve the performance of our approach. Experimental results demonstrate that our approach achieves superior performance at speeds that far exceed the frame-rate requirement on recent tracking benchmarks.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018EDL8074/_p
Copier
@ARTICLE{e101-d_11_2849,
author={Peng GAO, Yipeng MA, Chao LI, Ke SONG, Yan ZHANG, Fei WANG, Liyi XIAO, },
journal={IEICE TRANSACTIONS on Information},
title={Adaptive Object Tracking with Complementary Models},
year={2018},
volume={E101-D},
number={11},
pages={2849-2854},
abstract={Most state-of-the-art discriminative tracking approaches are based on either template appearance models or statistical appearance models. Despite template appearance models have shown excellent performance, they perform poorly when the target appearance changes rapidly. In contrast, statistic appearance models are insensitive to fast target state changes, but they yield inferior tracking results in challenging scenarios such as illumination variations and background clutters. In this paper, we propose an adaptive object tracking approach with complementary models based on template and statistical appearance models. Both of these models are unified via our novel combination strategy. In addition, we introduce an efficient update scheme to improve the performance of our approach. Experimental results demonstrate that our approach achieves superior performance at speeds that far exceed the frame-rate requirement on recent tracking benchmarks.},
keywords={},
doi={10.1587/transinf.2018EDL8074},
ISSN={1745-1361},
month={November},}
Copier
TY - JOUR
TI - Adaptive Object Tracking with Complementary Models
T2 - IEICE TRANSACTIONS on Information
SP - 2849
EP - 2854
AU - Peng GAO
AU - Yipeng MA
AU - Chao LI
AU - Ke SONG
AU - Yan ZHANG
AU - Fei WANG
AU - Liyi XIAO
PY - 2018
DO - 10.1587/transinf.2018EDL8074
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E101-D
IS - 11
JA - IEICE TRANSACTIONS on Information
Y1 - November 2018
AB - Most state-of-the-art discriminative tracking approaches are based on either template appearance models or statistical appearance models. Despite template appearance models have shown excellent performance, they perform poorly when the target appearance changes rapidly. In contrast, statistic appearance models are insensitive to fast target state changes, but they yield inferior tracking results in challenging scenarios such as illumination variations and background clutters. In this paper, we propose an adaptive object tracking approach with complementary models based on template and statistical appearance models. Both of these models are unified via our novel combination strategy. In addition, we introduce an efficient update scheme to improve the performance of our approach. Experimental results demonstrate that our approach achieves superior performance at speeds that far exceed the frame-rate requirement on recent tracking benchmarks.
ER -