Probabilistic Trajectory GOSPA: A Metric for Uncertainty-Aware Multi-Object Tracking Performance Evaluation

TL;DR

This paper introduces a probabilistic extension of the trajectory GOSPA metric for evaluating multi-object tracking algorithms that incorporate uncertainty, enabling comprehensive performance assessment including localization, existence, and track switching errors.

Contribution

It generalizes the trajectory GOSPA metric to account for uncertainties in object states and existence, maintaining polynomial-time computability and interpretability.

Findings

The metric effectively captures localization and existence uncertainties.

Decomposition yields intuitive cost components for various tracking errors.

Simulation results demonstrate its practical utility.

Abstract

This paper presents a generalization of the trajectory general optimal sub-pattern assignment (GOSPA) metric for evaluating multi-object tracking algorithms that provide trajectory estimates with track-level uncertainties. This metric builds on the recently introduced probabilistic GOSPA metric to account for both the existence and state estimation uncertainties of individual object states. Similar to trajectory GOSPA (TGOSPA), it can be formulated as a multidimensional assignment problem, and its linear programming relaxation--also a valid metric--is computable in polynomial time. Additionally, this metric retains the interpretability of TGOSPA, and we show that its decomposition yields intuitive costs terms associated to expected localization error and existence probability mismatch error for properly detected objects, expected missed and false detection error, and track switch error. The effectiveness of the proposed metric is demonstrated through a simulation study.