Glance-MCMT: A General MCMT Framework with Glance Initialization and Progressive Association

TL;DR

This paper introduces Glance-MCMT, a comprehensive multi-camera multi-target tracking framework that maintains consistent identities across views through a novel glance initialization and progressive association strategy, integrating trajectory and appearance cues.

Contribution

It presents a new MCMT tracking framework with a glance-based initialization and a prioritized global matching approach for improved identity consistency.

Findings

Effective global identity assignment across multiple views.

Reduced ID switches through prioritized matching.

Integration of 3D spatial validation enhances accuracy.

Abstract

We propose a multi-camera multi-target (MCMT) tracking framework that ensures consistent global identity assignment across views using trajectory and appearance cues. The pipeline starts with BoT-SORT-based single-camera tracking, followed by an initial glance phase to initialize global IDs via trajectory-feature matching. In later frames, new tracklets are matched to existing global identities through a prioritized global matching strategy. New global IDs are only introduced when no sufficiently similar trajectory or feature match is found. 3D positions are estimated using depth maps and calibration for spatial validation.