FusionTrack: End-to-End Multi-Object Tracking in Arbitrary Multi-View Environment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Xiaohe, Li, Pengfei, Fan, Zide, Geng, Ying, Mou, Fangli, Wu, Haohua, Ge, Yunping
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913857642954752
author Li, Xiaohe
Li, Pengfei
Fan, Zide
Geng, Ying
Mou, Fangli
Wu, Haohua
Ge, Yunping
author_facet Li, Xiaohe
Li, Pengfei
Fan, Zide
Geng, Ying
Mou, Fangli
Wu, Haohua
Ge, Yunping
contents Multi-view multi-object tracking (MVMOT) has found widespread applications in intelligent transportation, surveillance systems, and urban management. However, existing studies rarely address genuinely free-viewpoint MVMOT systems, which could significantly enhance the flexibility and scalability of cooperative tracking systems. To bridge this gap, we first construct the Multi-Drone Multi-Object Tracking (MDMOT) dataset, captured by mobile drone swarms across diverse real-world scenarios, initially establishing the first benchmark for multi-object tracking in arbitrary multi-view environment. Building upon this foundation, we propose \textbf{FusionTrack}, an end-to-end framework that reasonably integrates tracking and re-identification to leverage multi-view information for robust trajectory association. Extensive experiments on our MDMOT and other benchmark datasets demonstrate that FusionTrack achieves state-of-the-art performance in both single-view and multi-view tracking.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18727
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FusionTrack: End-to-End Multi-Object Tracking in Arbitrary Multi-View Environment
Li, Xiaohe
Li, Pengfei
Fan, Zide
Geng, Ying
Mou, Fangli
Wu, Haohua
Ge, Yunping
Computer Vision and Pattern Recognition
Multi-view multi-object tracking (MVMOT) has found widespread applications in intelligent transportation, surveillance systems, and urban management. However, existing studies rarely address genuinely free-viewpoint MVMOT systems, which could significantly enhance the flexibility and scalability of cooperative tracking systems. To bridge this gap, we first construct the Multi-Drone Multi-Object Tracking (MDMOT) dataset, captured by mobile drone swarms across diverse real-world scenarios, initially establishing the first benchmark for multi-object tracking in arbitrary multi-view environment. Building upon this foundation, we propose \textbf{FusionTrack}, an end-to-end framework that reasonably integrates tracking and re-identification to leverage multi-view information for robust trajectory association. Extensive experiments on our MDMOT and other benchmark datasets demonstrate that FusionTrack achieves state-of-the-art performance in both single-view and multi-view tracking.
title FusionTrack: End-to-End Multi-Object Tracking in Arbitrary Multi-View Environment
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.18727