Deep convolutional neural networks for thermal infrared object tracking

作者：

Highlights：

• The first contribution is transfering the pre-trained VGG-Net to thermal infrared tracking.

• The second contribution is proposing a correlation filter based ensemble tracker with multi-layer convolutional features.

• The third contribution is proposing a fusion method based Kullback–Leibler divergence.

摘要

•The first contribution is transfering the pre-trained VGG-Net to thermal infrared tracking.•The second contribution is proposing a correlation filter based ensemble tracker with multi-layer convolutional features.•The third contribution is proposing a fusion method based Kullback–Leibler divergence.

论文关键词：Thermal infrared tracking,Convolutional features,Correlation filter,Ensemble method

论文评审过程：Received 21 February 2017, Revised 22 July 2017, Accepted 25 July 2017, Available online 26 July 2017, Version of Record 13 September 2017.

论文官网地址：https://doi.org/10.1016/j.knosys.2017.07.032