Scholarlink

<p>Most two-stage detection and tracking methods currently adopt IoU distance and appearance features when matching target detection results with trajectory predictions. However, this approach has shortcomings in two main aspects: 1. The selection of feature extraction networks often relies on established methods, such as ReID-related methods, which are typically complex and significantly slow down algorithm processing speed. 2. Feature extraction networks lack consistency when extracting single-frame features, leading to difficulties in matching incomplete objects with previous features when occlusions occur, resulting in incorrect updates of target identities. To address these limitations, we propose a lightweight feature extraction network that integrates current target features with historical data. This network enhances data processing speed while obtaining smooth object features with temporal coherence. To address feature granularity variations across targets, we employ adaptive feature matching approaches tailored to different detection conditions. This strategy maximizes information efficiency while conserving computational resources. An adaptive multifeature extraction method is also proposed to accommodate diverse requirements across different scenarios. The matching strategy developed allows the tracker to serve as a universal and practical solution for various tracking scenarios within a unified framework, eliminating the need for manual parameter tuning. The effectiveness of PreAFTrack is demonstrated through comprehensive evaluations on three major MOT datasets: MOT17, MOT20, and DanceTrack. Code is available at <a href="https://github.com/Flowercaty/PreAFTrack">https://github.com/Flowercaty/PreAFTrack</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Preaftrack: multi-object tracking based on adaptive feature matching from detection results

  • 期刊论文

  • 发表日期：

    2025年7月7日

查看全文
  • Mandun Zhang,
  • Chenyue Hou,
  • Yonghui Pang,
  • Jiale Shi,
  • Yiqi Zhang,
  • Xiangsheng Huang

摘要

Most two-stage detection and tracking methods currently adopt IoU distance and appearance features when matching target detection results with trajectory predictions. However, this approach has shortcomings in two main aspects: 1. The selection of feature extraction networks often relies on established methods, such as ReID-related methods, which are typically complex and significantly slow down algorithm processing speed. 2. Feature extraction networks lack consistency when extracting single-frame features, leading to difficulties in matching incomplete objects with previous features when occlusions occur, resulting in incorrect updates of target identities. To address these limitations, we propose a lightweight feature extraction network that integrates current target features with historical data. This network enhances data processing speed while obtaining smooth object features with temporal coherence. To address feature granularity variations across targets, we employ adaptive feature matching approaches tailored to different detection conditions. This strategy maximizes information efficiency while conserving computational resources. An adaptive multifeature extraction method is also proposed to accommodate diverse requirements across different scenarios. The matching strategy developed allows the tracker to serve as a universal and practical solution for various tracking scenarios within a unified framework, eliminating the need for manual parameter tuning. The effectiveness of PreAFTrack is demonstrated through comprehensive evaluations on three major MOT datasets: MOT17, MOT20, and DanceTrack. Code is available at https://github.com/Flowercaty/PreAFTrack.