HCDC-SRCF tracker: Learning an adaptively multi-feature fuse tracker in spatial regularized correlation filters framework

作者：

Highlights：

• Firstly, our HCDC-SRCF tracker simultaneously makes the best of complementary advantages from multi-layer DC features and HC features.

• Secondly, the alternating direction algorithm of multipliers (ADMM) is introduced into spatially regularized correlation filters to address the problem on how to search target object.

• Finally, we demonstrate the effectiveness, accuracy, and robustness of the proposed tracker on five different benchmark datasets.

摘要

•Firstly, our HCDC-SRCF tracker simultaneously makes the best of complementary advantages from multi-layer DC features and HC features.•Secondly, the alternating direction algorithm of multipliers (ADMM) is introduced into spatially regularized correlation filters to address the problem on how to search target object.•Finally, we demonstrate the effectiveness, accuracy, and robustness of the proposed tracker on five different benchmark datasets.

论文关键词：Multi-layer convolutional features,Hand-crafted features,Visual object tracking,Convolutional neural network

论文评审过程：Received 12 September 2021, Revised 25 November 2021, Accepted 5 December 2021, Available online 11 December 2021, Version of Record 24 December 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107913