Visual object tracking with adaptive structural convolutional network

作者：

Highlights：

• Propose an adaptive structural convolutional filter network for visual tracking.

• The structural filter layer can capture the target’s structural patterns.

• Develop an adaptive weighting strategy to improve the tracker’s stability.

• Experimental results demonstrate the efficiency of the proposed tracker.

摘要

•Propose an adaptive structural convolutional filter network for visual tracking.•The structural filter layer can capture the target’s structural patterns.•Develop an adaptive weighting strategy to improve the tracker’s stability.•Experimental results demonstrate the efficiency of the proposed tracker.

论文关键词：Visual tracking,Convolution neural network,Structural filters,Adaptive weighting

论文评审过程：Received 2 October 2019, Revised 20 January 2020, Accepted 22 January 2020, Available online 24 January 2020, Version of Record 18 May 2020.

论文官网地址：https://doi.org/10.1016/j.knosys.2020.105554