[TPAMI-2023] Effective Local and Global Search for Fast Long-Term Tracking
论文阅读 [TPAMI-2023] Effective Local and Global Search for Fast Long-Term Tracking
论文搜索(studyai.com)
搜索论文: Effective Local and Global Search for Fast Long-Term Tracking
搜索论文: http://www.studyai.com/search/whole-site/?q=Effective+Local+and+Global+Search+for+Fast+Long-Term+Tracking&fr=csdn
关键字(Keywords)
Target tracking; Search problems; Task analysis; Proposals; Object tracking; Correlation; Benchmark testing; Visual object tracking; long-term tracking; global re-detection
机器视觉
视觉跟踪
摘要(Abstract)
Compared with short-term tracking, long-term tracking remains a challenging task that usually requires the tracking algorithm to track targets within a local region and re-detect targets over the entire image.
与短期跟踪相比，长期跟踪仍然是一项具有挑战性的任务，通常需要跟踪算法在局部区域内跟踪目标并在整个图像上重新检测目标。
However, few works have been done and their performances have also been limited.
然而，已经完成的工作很少，其性能也受到限制。
In this paper, we present a novel robust and real-time long-term tracking framework based on the proposed local search module and re-detection module.
在本文中，我们基于所提出的局部搜索模块和重新检测模块，提出了一种新的鲁棒实时长期跟踪框架。
The local search module consists of an effective bounding box regressor to generate a series of candidate proposals and a target verifier to infer the optimal candidate with its confidence score.
本地搜索模块由一个有效的边界框回归器和一个目标验证器组成，前者用于生成一系列候选方案，后者用于推断具有置信度的最佳候选方案。
For local search, we design a long short-term updated scheme to improve the target verifier.
对于本地搜索，我们设计了一个长期-短期更新方案来改进目标验证器。
The verification capability of the tracker can be improved by using several templates updated at different times.
通过使用在不同时间更新的几个模板，可以提高跟踪器的验证能力。
Based on the verification scores, our tracker determines whether the tracked object is present or absent and then chooses the tracking strategies of local or global search, respectively, in the next frame.
基于验证分数，我们的跟踪器确定被跟踪对象是否存在，然后在下一帧中分别选择本地或全局搜索的跟踪策略。
For global re-detection, we develop a novel re-detection module that can estimate the target position and target size for a given base tracker.
对于全局重新检测，我们开发了一种新的重新检测模块，可以估计给定基础跟踪器的目标位置和目标大小。
We conduct a series of experiments to demonstrate that this module can be flexibly integrated into many other tracking algorithms for long-term tracking and that it can improve long-term tracking performance effectively.
我们进行了一系列实验，以证明该模块可以灵活地集成到许多其他跟踪算法中进行长期跟踪，并且可以有效地提高长期跟踪性能。
Numerous experiments and discussions are conducted on several popular tracking datasets, including VOT, OxUvA, TLP, and LaSOT.
在几个流行的跟踪数据集上进行了大量实验和讨论，包括VOT、OxUvA、TLP和LaSOT。
The experimental results demonstrate that the proposed tracker achieves satisfactory performance with a real-time speed.
实验结果表明，所提出的跟踪器以实时速度实现了令人满意的性能。
Code is available at https://github.com/difhnp/ELGLT…
代码位于https://github.com/difhnp/ELGLT.。
作者(Authors)
[‘Haojie Zhao’, ‘Bin Yan’, ‘Dong Wang’, ‘Xuesheng Qian’, ‘Xiaoyun Yang’, ‘Huchuan Lu’]