Trajectory Poisson multi-Bernoulli filters
This work addresses multi-target tracking problems, likely for applications like surveillance or robotics, by offering computationally lighter alternatives to existing filters, representing an incremental improvement.
The paper tackles multi-target tracking by developing two trajectory Poisson multi-Bernoulli (TPMB) filters to estimate sets of alive or all trajectories, based on propagating a PMB density and using Kullback-Leibler divergence minimization to maintain computational efficiency. The filters are shown to outperform previous multi-target tracking algorithms, though no specific numerical results are provided in the abstract.
This paper presents two trajectory Poisson multi-Bernoulli (TPMB) filters for multi-target tracking: one to estimate the set of alive trajectories at each time step and another to estimate the set of all trajectories, which includes alive and dead trajectories, at each time step. The filters are based on propagating a Poisson multi-Bernoulli (PMB) density on the corresponding set of trajectories through the filtering recursion. After the update step, the posterior is a PMB mixture (PMBM) so, in order to obtain a PMB density, a Kullback-Leibler divergence minimisation on an augmented space is performed. The developed filters are computationally lighter alternatives to the trajectory PMBM filters, which provide the closed-form recursion for sets of trajectories with Poisson birth model, and are shown to outperform previous multi-target tracking algorithms.