How Can Subsampling Reduce Complexity in Sequential MCMC Methods and Deal with Big Data in Target Tracking?

TL;DR

This paper introduces an adaptive subsampling method for sequential MCMC in target tracking, significantly reducing computational complexity while maintaining accuracy in large data environments.

Contribution

It proposes a novel adaptive subsampling technique for sequential MCMC that effectively balances data volume and tracking accuracy in cluttered scenarios.

Findings

Over 40% reduction in processing time

Negligible loss in tracking accuracy

Effective handling of large data volumes in real-time

Abstract

Target tracking faces the challenge in coping with large volumes of data which requires efficient methods for real time applications. The complexity considered in this paper is when there is a large number of measurements which are required to be processed at each time step. Sequential Markov chain Monte Carlo (MCMC) has been shown to be a promising approach to target tracking in complex environments, especially when dealing with clutter. However, a large number of measurements usually results in large processing requirements. This paper goes beyond the current state-of-the-art and presents a novel Sequential MCMC approach that can overcome this challenge through adaptively subsampling the set of measurements. Instead of using the whole large volume of available data, the proposed algorithm performs a trade off between the number of measurements to be used and the desired accuracy of the estimates to be obtained in the presence of clutter. We show results with large improvements in processing time, more than 40% with a negligible loss in tracking performance, compared with the solution without subsampling.