Particle filtering with invertible particle flow

Abstract

A key challenge when designing particle filters in high-dimensional state spaces is the construction of a proposal distribution that is close to the posterior distribution. Recent advances in particle flow filters provide a promising avenue to avoid weight degeneracy; particles drawn from the prior distribution are migrated in the state-space to the posterior distribution by solving partial differential equations.

Numerous particle flow filters have been proposed based on different assumptions concerning the flow dynamics. Approximations are needed in the implementation of all of these filters; as a result the particles do not exactly match a sample drawn from the desired posterior distribution. Past efforts to correct the discrepancies involve expensive calculations of importance weights.

In this talk, I will present new filters which incorporate deterministic particle flows into an encompassing particle filter framework. The valuable theoretical guarantees concerning particle filter performance still apply, but we can exploit the attractive performance of the particle flow methods.

The filters we describe involve a computationally efficient weight update step, arising because the embedded particle flows we design possess an invertible mapping property. We evaluate the proposed particle flow particle filters' performance through numerical simulations of a challenging multi-target multi-sensor tracking scenario and complex high-dimensional filtering examples.

Biography

Yunpeng Li received the Ph.D. degree in electrical engineering from McGill University, Montreal, Canada, in 2017, after receiving the B.A. and MEng. degrees from the Beijing University of Posts and Telecommunications, Beijing, China, in 2009 and 2012, respectively.

He has been with the Machine Learning Research Group, University of Oxford, Oxford, UK since April 2017, where he is a Postdoctoral Research Assistant in Machine Learning. He is a Junior Research Fellow at the Wolfson College, Oxford.

His research interests include:

• Bayesian and Monte Carlo inference in high-dimensional spaces
• Radio-frequency tomographic tracking
• Microwave breast cancer detection
• Environmental acoustic sensing.

His current research on detecting malaria-vectoring mosquitoes with low-cost mobile phones has been covered by MIT Technology Review, Digital Trends, New Scientist, The Guardian, among other venues.