Common tracking systems hide a hidden flaw that makes advanced filtering useless
Researchers have identified a class of tracking problems where a supposedly superior filtering algorithm produces no better results than a basic alternative—but engineers have no easy way to detect it beforehand. The finding matters because companies and agencies relying on these systems for autonomous vehicles, robotics, and surveillance may be wasting computational resources without realizing it.
Originaltitel: Marginalized Particle Filter Degeneration
<p>The marginalized particle filter (MPF) is known to often outperform the standard particle filter (PF) in terms of estimation accuracy for the same number of particles. This is due to the fact that the MPF uses a Kalman filter (KF) to handle a linear sub-part of the system in a more efficient way than a particle approximation. However, for certain systems, marginalization implies no gain (MING), resulting in identical output distribution, including estimation accuracy. This article explains when and why this happens in two steps. The first step is to show that the MPF and PF algorithms are algebraically identical in the case the KF covariance is zero. The second step is to show that for MING systems the KF in the MPF degenerates, in that the KF covariance approaches zero. We provide a practical guide for easily checking if a system is a MING system or not, and show by examples that MING systems exist, and in fact are quite common in, e.g., tracking applications.</p>