We combine the siamese network and the recurrent regression network, proposing a two-stage tracking framework termed as SiamReg. Our method solves the problem that the classic siamese network can not judge the target size precisely and simplifies the procedures of regression in the training and testing process. We perform experiments on three challenging tracking datasets: VOT2016, OTB100, and VOT2018. The results indicate that, after offline trained, SiamReg can obtain a higher expected average overlap measure.
Yao GE
Nanjing University of Posts and Telecommunications
Rui CHEN
Nanjing Institute of Technology
Ying TONG
Nanjing Institute of Technology
Xuehong CAO
Nanjing Institute of Technology
Ruiyu LIANG
Nanjing 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.
Copy
Yao GE, Rui CHEN, Ying TONG, Xuehong CAO, Ruiyu LIANG, "Combining Siamese Network and Regression Network for Visual Tracking" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 8, pp. 1924-1927, August 2020, doi: 10.1587/transinf.2020EDL8032.
Abstract: We combine the siamese network and the recurrent regression network, proposing a two-stage tracking framework termed as SiamReg. Our method solves the problem that the classic siamese network can not judge the target size precisely and simplifies the procedures of regression in the training and testing process. We perform experiments on three challenging tracking datasets: VOT2016, OTB100, and VOT2018. The results indicate that, after offline trained, SiamReg can obtain a higher expected average overlap measure.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2020EDL8032/_p
Copy
@ARTICLE{e103-d_8_1924,
author={Yao GE, Rui CHEN, Ying TONG, Xuehong CAO, Ruiyu LIANG, },
journal={IEICE TRANSACTIONS on Information},
title={Combining Siamese Network and Regression Network for Visual Tracking},
year={2020},
volume={E103-D},
number={8},
pages={1924-1927},
abstract={We combine the siamese network and the recurrent regression network, proposing a two-stage tracking framework termed as SiamReg. Our method solves the problem that the classic siamese network can not judge the target size precisely and simplifies the procedures of regression in the training and testing process. We perform experiments on three challenging tracking datasets: VOT2016, OTB100, and VOT2018. The results indicate that, after offline trained, SiamReg can obtain a higher expected average overlap measure.},
keywords={},
doi={10.1587/transinf.2020EDL8032},
ISSN={1745-1361},
month={August},}
Copy
TY - JOUR
TI - Combining Siamese Network and Regression Network for Visual Tracking
T2 - IEICE TRANSACTIONS on Information
SP - 1924
EP - 1927
AU - Yao GE
AU - Rui CHEN
AU - Ying TONG
AU - Xuehong CAO
AU - Ruiyu LIANG
PY - 2020
DO - 10.1587/transinf.2020EDL8032
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2020
AB - We combine the siamese network and the recurrent regression network, proposing a two-stage tracking framework termed as SiamReg. Our method solves the problem that the classic siamese network can not judge the target size precisely and simplifies the procedures of regression in the training and testing process. We perform experiments on three challenging tracking datasets: VOT2016, OTB100, and VOT2018. The results indicate that, after offline trained, SiamReg can obtain a higher expected average overlap measure.
ER -