On-line fusion of trackers for single-object tracking

作者：

Highlights：

• The work focuses on the design of good strategies for the on-line fusion of trackers.

• Fusion can operate at two levels: tracker output selection and model correction.

• We show experimentally that the ability to predict drift is essential for fusion.

• We propose a tracker complementarity measure to choose the best tracker combination.

摘要

•The work focuses on the design of good strategies for the on-line fusion of trackers.•Fusion can operate at two levels: tracker output selection and model correction.•We show experimentally that the ability to predict drift is essential for fusion.•We propose a tracker complementarity measure to choose the best tracker combination.

论文关键词：Visual object tracking,Fusion,On-line behavior analysis

论文评审过程：Received 21 April 2016, Revised 15 September 2017, Accepted 16 September 2017, Available online 19 September 2017, Version of Record 5 October 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.09.026