LTST: Long-term segmentation tracker with memory attention network

作者：

Highlights：

• Propose a framework LTST for both long-term tracking and object segmentation.

• Propose a memory attention network to leverage historical information.

• Supplementary modules are integrated for tracking verification and re-detection.

• Experiments demonstrate that LTST achieves state-of-the art performance.

摘要

Highlights•Propose a framework LTST for both long-term tracking and object segmentation.•Propose a memory attention network to leverage historical information.•Supplementary modules are integrated for tracking verification and re-detection.•Experiments demonstrate that LTST achieves state-of-the art performance.

论文关键词：Long-term tracking,Object segmentation,Memory network,Attention mechanism

论文评审过程：Received 1 July 2021, Revised 28 October 2021, Accepted 10 January 2022, Available online 14 January 2022, Version of Record 25 January 2022.

论文官网地址：https://doi.org/10.1016/j.imavis.2022.104374