A time-weighted metric for sets of trajectories to assess multi-object tracking algorithms

TL;DR

This paper introduces a novel time-weighted metric for evaluating multi-object tracking algorithms, allowing for flexible, application-specific assessments of tracking accuracy over time.

Contribution

It extends existing metrics by incorporating time-dependent weights and provides a computationally feasible linear programming approach.

Findings

The metric is a true metric and can be computed in polynomial time.

It can be extended to evaluate random finite sets of trajectories.

The approach offers greater flexibility for different tracking scenarios.

Abstract

This paper proposes a metric for sets of trajectories to evaluate multi-object tracking algorithms that includes time-weighted costs for localisation errors of properly detected targets, for false targets, missed targets and track switches. The proposed metric extends the metric in [1] by including weights to the costs associated to different time steps. The time-weighted costs increase the flexibility of the metric [1] to fit more applications and user preferences. We first introduce a metric based on multi-dimensional assignments, and then its linear programming relaxation, which is computable in polynomial time and is also a metric. The metrics can also be extended to metrics on random finite sets of trajectories to evaluate and rank algorithms across different scenarios, each with a ground truth set of trajectories.