Single-cluster PHD filter methods for joint multi-object filtering and parameter estimation

Many multi-object estimation problems require additional estimation of model or sensor parameters that are either common to all objects or related to unknown characterisation of one or more sensors. Important examples of these include registration of multiple sensors, estimating clutter profiles, and robot localisation. Often these parameters are estimated separately to the multi-object estimation process, which can lead to systematic errors or overconfidence in the estimates. These parameters can be estimated jointly with the multi-object process based only on the sensor data using a single-cluster point process model. This paper presents novel results for joint parameter estimation and multi-object filtering based on a single-cluster second-order Probability Hypothesis Density (PHD) and Cardinalized PHD (CPHD) filter. Experiments provide a comparison between the discussed approaches using different likelihood functions.