Multi-object Tracking for Generic Observation Model Using Labeled Random Finite Sets
For researchers in multi-object tracking, this work provides a principled framework that handles generic observation models without simplifying assumptions, but the novelty is incremental as it extends existing labeled RFS theory.
This paper proposes an exact Bayesian filtering solution for multi-object tracking under generic observation models using labeled random finite sets, and develops computationally tractable approximations via Kullback-Leibler divergence minimization and dynamic grouping. Numerical experiments demonstrate improved tracking accuracy and computational efficiency compared to state-of-the-art methods.
This paper presents an exact Bayesian filtering solution for the multi-object tracking problem with the generic observation model. The proposed solution is designed in the labeled random finite set framework, using the product styled representation of labeled multi-object densities, with the standard multi-object transition kernel and no particular simplifying assumptions on the multi-object likelihood. Computationally tractable solutions are also devised by applying a principled approximation involving the replacement of the full multi-object density with a labeled multi-Bernoulli density that minimizes the Kullback-Leibler divergence and preserves the first-order moment. To achieve the fast performance, a dynamic grouping procedure based implementation is presented with a step-by-step algorithm. The performance of the proposed filter and its tractable implementations are verified and compared with the state-of-the-art in numerical experiments.