Tracklet Association by Online Target-Specific Metric Learning and Coherent Dynamics Estimation

TL;DR

This paper introduces a novel online multi-person tracking method that learns target-specific appearance and motion cues to improve long-term tracking accuracy, especially during occlusions and interactions.

Contribution

It proposes a new framework combining online target-specific metric learning with coherent dynamics estimation for improved tracklet association.

Findings

Outperforms several state-of-the-art tracking methods on public datasets.

Effectively handles occlusions and object interactions.

Learns cue weights online to adapt to difficult tracking scenarios.

Abstract

In this paper, we present a novel method based on online target-specific metric learning and coherent dynamics estimation for tracklet (track fragment) association by network flow optimization in long-term multi-person tracking. Our proposed framework aims to exploit appearance and motion cues to prevent identity switches during tracking and to recover missed detections. Furthermore, target-specific metrics (appearance cue) and motion dynamics (motion cue) are proposed to be learned and estimated online, i.e. during the tracking process. Our approach is effective even when such cues fail to identify or follow the target due to occlusions or object-to-object interactions. We also propose to learn the weights of these two tracking cues to handle the difficult situations, such as severe occlusions and object-to-object interactions effectively. Our method has been validated on several public datasets and the experimental results show that it outperforms several state-of-the-art tracking methods.