Answers are grounded in this paper and cite their sources.
This work extends the extended target cluster process tracker [1] on learned spatial distribution by a first-order approximation in its associations. The first-order approximation provides a uni-modal posterior of the closed-form Bayesian recursion and thus, significant improvements in computational complexity and runtime at relatively little cost in estimation quality. We compare the relative performance of the proposed model with previously proposed approaches in a large-scale evaluation based on the Nuscenes data set.
No takes yet. Share an insight, caveat, or question.
Honer et al. (2023) studied this question.