Learning region sparse constraint correlation filter for tracking
作者：
Highlights：
• A more effective region sparse constraint correlation filter is proposed for interference rejection.
• A binary mask is used for the sparsity of the target region instead of the whole sample region.
• A context-aware term is integrated to enhance the discriminant ability of the filter.
• An ADMM optimization algorithm is proposed to solve the model.
• Performance analyses and experimental validation show that the proposed method indeed works.
摘要
•A more effective region sparse constraint correlation filter is proposed for interference rejection.•A binary mask is used for the sparsity of the target region instead of the whole sample region.•A context-aware term is integrated to enhance the discriminant ability of the filter.•An ADMM optimization algorithm is proposed to solve the model.•Performance analyses and experimental validation show that the proposed method indeed works.
论文关键词：Correlation filters,Visual tracking,Elastic net regression
论文评审过程：Received 17 March 2020, Revised 12 September 2020, Accepted 21 October 2020, Available online 1 November 2020, Version of Record 9 November 2020.
论文官网地址：https://doi.org/10.1016/j.image.2020.116042
