Computer Model for Evaluating Multi-Target Tracking Algorithms

Garret Vo, Chiwoo Park
{"title":"Computer Model for Evaluating Multi-Target Tracking Algorithms","authors":"Garret Vo, Chiwoo Park","doi":"10.4236/OJMSI.2019.71001","DOIUrl":null,"url":null,"abstract":"Public benchmark datasets have been widely used to evaluate multi-target tracking algorithms. Ideally, the benchmark datasets should include the video scenes of all scenarios that need to be tested. However, a limited amount of the currently available benchmark datasets does not comprehensively cover all necessary test scenarios. This limits the evaluation of multitarget tracking algorithms with various test scenarios. This paper introduced a computer simulation model that generates benchmark datasets for evaluating multi-target tracking algorithms with the complexity of multitarget tracking scenarios directly controlled by simulation inputs such as target birth and death rates, target movement, the rates of target merges and splits, target appearances, and image noise types and levels. The simulation model generated a simulated video and also provides the ground-truth target tracking for the simulated video, so the evaluation of multitarget tracking algorithms can be easily performed without any manual video annotation process. We demonstrated the use of the proposed simulation model for evaluating tracking-by-detection algorithms and filtering-based tracking algorithms.","PeriodicalId":56990,"journal":{"name":"建模与仿真(英文)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2019-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"建模与仿真(英文)","FirstCategoryId":"1093","ListUrlMain":"https://doi.org/10.4236/OJMSI.2019.71001","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Public benchmark datasets have been widely used to evaluate multi-target tracking algorithms. Ideally, the benchmark datasets should include the video scenes of all scenarios that need to be tested. However, a limited amount of the currently available benchmark datasets does not comprehensively cover all necessary test scenarios. This limits the evaluation of multitarget tracking algorithms with various test scenarios. This paper introduced a computer simulation model that generates benchmark datasets for evaluating multi-target tracking algorithms with the complexity of multitarget tracking scenarios directly controlled by simulation inputs such as target birth and death rates, target movement, the rates of target merges and splits, target appearances, and image noise types and levels. The simulation model generated a simulated video and also provides the ground-truth target tracking for the simulated video, so the evaluation of multitarget tracking algorithms can be easily performed without any manual video annotation process. We demonstrated the use of the proposed simulation model for evaluating tracking-by-detection algorithms and filtering-based tracking algorithms.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
评价多目标跟踪算法的计算机模型
公共基准数据集已被广泛用于评估多目标跟踪算法。理想情况下,基准数据集应该包括需要测试的所有场景的视频场景。然而,目前可用的基准数据集数量有限,不能全面覆盖所有必要的测试场景。这限制了多目标跟踪算法在各种测试场景下的评估。本文介绍了一种计算机仿真模型,该模型生成基准数据集,用于评估多目标跟踪算法,多目标跟踪场景的复杂性由仿真输入直接控制,如目标的出生率和死亡率、目标的运动、目标的合并和分裂率、目标的外观以及图像噪声的类型和水平。该仿真模型生成了仿真视频,并为仿真视频提供了真实目标跟踪,因此无需手动视频注释过程即可方便地对多目标跟踪算法进行评估。我们演示了使用所提出的仿真模型来评估检测跟踪算法和基于滤波的跟踪算法。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
自引率
0.00%
发文量
61
期刊最新文献
Comparative Evaluation of the Performance of SWAT, SWAT+, and APEX Models in Simulating Edge of Field Hydrological Processes Making Sense of Anything thru Analytics: Employees Provident Fund (EPF) Simulation of Crack Pattern Formation Due to Shrinkage in a Drying Material Modelling COVID-19 Cumulative Number of Cases in Kenya Using a Negative Binomial INAR (1) Model Understanding the Dynamics Location of Very Large Populations Interacted with Service Points
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1