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

TL;DR

FusionTrack is an end-to-end multi-object tracking framework that leverages multi-view data in arbitrary environments, supported by a new drone-based dataset and benchmark, achieving state-of-the-art results.

Contribution

The paper introduces FusionTrack, a novel end-to-end multi-view tracking method, and provides the first multi-drone dataset and benchmark for arbitrary multi-view environments.

Findings

FusionTrack outperforms existing methods on multiple datasets.

The new MDMOT dataset enables robust evaluation in real-world scenarios.

FusionTrack effectively integrates tracking and re-identification for improved accuracy.

Abstract

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.