Probabilistic GOSPA: A Metric for Performance Evaluation of Multi-Object Filters with Uncertainties

TL;DR

This paper introduces P-GOSPA, a probabilistic extension of the GOSPA metric, designed to evaluate multi-object filters by accounting for uncertainties in probabilistic multi-object representations.

Contribution

The paper develops P-GOSPA, extending GOSPA into the space of multi-Bernoulli densities while preserving interpretability and error decomposition.

Findings

P-GOSPA effectively evaluates multi-object filters with uncertainties.

Simulations demonstrate P-GOSPA's accuracy and interpretability.

P-GOSPA maintains GOSPA's error decomposition in probabilistic contexts.

Abstract

This paper presents a probabilistic generalization of the Generalized Optimal Sub-Pattern Assignment (GOSPA) metric, termed P-GOSPA. The GOSPA metric has been widely used to evaluate the distance between finite sets, particularly in multi-object estimation applications. The P-GOSPA extends GOSPA into the space of multi-Bernoulli densities, incorporating inherent uncertainty in probabilistic multi-object representations. Additionally, P-GOSPA retains the interpretability of GOSPA, such as its decomposition into localization, missed detection, and false detection errors in a sound and meaningful manner. Examples and simulations are provided to demonstrate the efficacy of the proposed P-GOSPA metric.