Re-identification framework for long term visual object tracking based on object detection and classification

作者：

Highlights：

• Novel long-term, modular tracking framework based on classification.

• Standard object detectors combined with trackers and control with a lightweight classifier.

• Real-time execution even on embedded devices.

• Experimental results on five challenging tracking benchmarks.

摘要

•Novel long-term, modular tracking framework based on classification.•Standard object detectors combined with trackers and control with a lightweight classifier.•Real-time execution even on embedded devices.•Experimental results on five challenging tracking benchmarks.

论文关键词：Visual object tracking,Long-term tracking,Re-detection,Deep learning

论文评审过程：Received 20 November 2019, Revised 12 May 2020, Accepted 2 August 2020, Available online 11 August 2020, Version of Record 12 August 2020.

论文官网地址：https://doi.org/10.1016/j.image.2020.115969