Variational PMB filter via coordinate descent Kullback-Leibler divergence minimisation
For multi-target tracking researchers, this provides a principled derivation and performance advantage for the V-PMB filter in challenging scenarios.
The paper derives the variational Poisson multi-Bernoulli (V-PMB) filter via coordinate descent Kullback-Leibler divergence minimisation on an augmented space, showing it preserves the probability hypothesis density. Comparisons demonstrate benefits over other PMB filters when targets come close and separate.
This paper presents a new derivation of the variational Poisson multi-Bernoulli (V-PMB) filter for multi-target estimation proposed in [#Williams15]. The proposed derivation is based on considering an augmented space that includes the set of target states with their track indices and the global hypothesis variable. Then, we show that the V-PMB projection performs a coordinate descent Kullback-Leibler divergence (KLD) minimisation on this augmented space to fit the best possible PMB density to the Poisson multi-Bernoulli mixture (PMBM) posterior. We also show that this V-PMB projection keeps the probability hypothesis density of the posterior. The paper also includes a comparison with the PMBM filter and other PMB filter variants, including a track-oriented Murty-based implementation, a track-oriented loopy belief propagation implementation and a global nearest neighbour implementation, showing the benefits of the V-PMB filter compared to the other PMB filters when targets get in close proximity and then separate.