Visual object tracking is always a challenging task in computer vision. During the tracking, the shape and appearance of the target may change greatly, and because of the lack of sufficient training samples, most of the online learning tracking algorithms will have performance bottlenecks. In this paper, an improved real-time algorithm based on deep learning features is proposed, which combines multi-feature fusion, multi-scale estimation, adaptive updating of target model and re-detection after target loss. The effectiveness and advantages of the proposed algorithm are proved by a large number of comparative experiments with other excellent algorithms on large benchmark datasets.
Xianyu WANG
Xidian University,Academy of Space Electronic Information Technology
Cong LI
Academy of Space Electronic Information Technology
Heyi LI
Xidian University
Rui ZHANG
Shaanxi Aerospace Technology Application Research Institute Co., Ltd.
Zhifeng LIANG
Shaanxi Aerospace Technology Application Research Institute Co., Ltd.
Hai WANG
Xidian University
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.
Copy
Xianyu WANG, Cong LI, Heyi LI, Rui ZHANG, Zhifeng LIANG, Hai WANG, "An Improved Real-Time Object Tracking Algorithm Based on Deep Learning Features" in IEICE TRANSACTIONS on Information,
vol. E106-D, no. 5, pp. 786-793, May 2023, doi: 10.1587/transinf.2022DLP0039.
Abstract: Visual object tracking is always a challenging task in computer vision. During the tracking, the shape and appearance of the target may change greatly, and because of the lack of sufficient training samples, most of the online learning tracking algorithms will have performance bottlenecks. In this paper, an improved real-time algorithm based on deep learning features is proposed, which combines multi-feature fusion, multi-scale estimation, adaptive updating of target model and re-detection after target loss. The effectiveness and advantages of the proposed algorithm are proved by a large number of comparative experiments with other excellent algorithms on large benchmark datasets.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2022DLP0039/_p
Copy
@ARTICLE{e106-d_5_786,
author={Xianyu WANG, Cong LI, Heyi LI, Rui ZHANG, Zhifeng LIANG, Hai WANG, },
journal={IEICE TRANSACTIONS on Information},
title={An Improved Real-Time Object Tracking Algorithm Based on Deep Learning Features},
year={2023},
volume={E106-D},
number={5},
pages={786-793},
abstract={Visual object tracking is always a challenging task in computer vision. During the tracking, the shape and appearance of the target may change greatly, and because of the lack of sufficient training samples, most of the online learning tracking algorithms will have performance bottlenecks. In this paper, an improved real-time algorithm based on deep learning features is proposed, which combines multi-feature fusion, multi-scale estimation, adaptive updating of target model and re-detection after target loss. The effectiveness and advantages of the proposed algorithm are proved by a large number of comparative experiments with other excellent algorithms on large benchmark datasets.},
keywords={},
doi={10.1587/transinf.2022DLP0039},
ISSN={1745-1361},
month={May},}
Copy
TY - JOUR
TI - An Improved Real-Time Object Tracking Algorithm Based on Deep Learning Features
T2 - IEICE TRANSACTIONS on Information
SP - 786
EP - 793
AU - Xianyu WANG
AU - Cong LI
AU - Heyi LI
AU - Rui ZHANG
AU - Zhifeng LIANG
AU - Hai WANG
PY - 2023
DO - 10.1587/transinf.2022DLP0039
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E106-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 2023
AB - Visual object tracking is always a challenging task in computer vision. During the tracking, the shape and appearance of the target may change greatly, and because of the lack of sufficient training samples, most of the online learning tracking algorithms will have performance bottlenecks. In this paper, an improved real-time algorithm based on deep learning features is proposed, which combines multi-feature fusion, multi-scale estimation, adaptive updating of target model and re-detection after target loss. The effectiveness and advantages of the proposed algorithm are proved by a large number of comparative experiments with other excellent algorithms on large benchmark datasets.
ER -