TY - GEN
T1 - Tracker-Level Decision by Deep Reinforcement Learning for Robust Visual Tracking
AU - Huang, Wenju
AU - Wu, Yuwei
AU - Jia, Yunde
N1 - Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - In this paper, we formulate the multi-tracker tracking problem as a decision-making task and train an expert by the deep reinforcement learning (DRL) to select the best tracker. Specifically, the expert takes the response map of the tracker as input and outputs a scalar to indicate the reliability of the tracker. With the DRL, the expert can make full use of complementary information among base trackers. Furthermore, under the guidance of the deep expert, base trackers update themselves adaptively to capture the changes of object appearance and prevent corruption. The experimental results on public tracking benchmarks demonstrate that the proposed method outperforms the state-of-the-art methods.
AB - In this paper, we formulate the multi-tracker tracking problem as a decision-making task and train an expert by the deep reinforcement learning (DRL) to select the best tracker. Specifically, the expert takes the response map of the tracker as input and outputs a scalar to indicate the reliability of the tracker. With the DRL, the expert can make full use of complementary information among base trackers. Furthermore, under the guidance of the deep expert, base trackers update themselves adaptively to capture the changes of object appearance and prevent corruption. The experimental results on public tracking benchmarks demonstrate that the proposed method outperforms the state-of-the-art methods.
KW - Deep reinforcement learning
KW - Tracker selection
KW - Visual object tracking
UR - https://www.scopus.com/pages/publications/85076921794
U2 - 10.1007/978-3-030-34120-6_36
DO - 10.1007/978-3-030-34120-6_36
M3 - Conference contribution
AN - SCOPUS:85076921794
SN - 9783030341190
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 442
EP - 453
BT - Image and Graphics - 10th International Conference, ICIG 2019, Proceedings, Part 1
A2 - Zhao, Yao
A2 - Lin, Chunyu
A2 - Barnes, Nick
A2 - Chen, Baoquan
A2 - Westermann, Rüdiger
A2 - Kong, Xiangwei
PB - Springer
T2 - 10th International Conference on Image and Graphics, ICIG 2019
Y2 - 23 August 2019 through 25 August 2019
ER -