An overview of differentiable particle filters for data-adaptive sequential Bayesian inference
This is an incremental overview paper that synthesizes recent advances in data-adaptive particle filtering for researchers in sequential estimation and machine learning.
The paper reviews differentiable particle filters, which use neural networks to adaptively optimize components like dynamic and measurement models for sequential Bayesian inference, highlighting their potential in complex tasks such as vision-based robot localization.
By approximating posterior distributions with weighted samples, particle filters (PFs) provide an efficient mechanism for solving non-linear sequential state estimation problems. While the effectiveness of particle filters has been recognised in various applications, their performance relies on the knowledge of dynamic models and measurement models, as well as the construction of effective proposal distributions. An emerging trend involves constructing components of particle filters using neural networks and optimising them by gradient descent, and such data-adaptive particle filtering approaches are often called differentiable particle filters. Due to the expressiveness of neural networks, differentiable particle filters are a promising computational tool for performing inference on sequential data in complex, high-dimensional tasks, such as vision-based robot localisation. In this paper, we review recent advances in differentiable particle filters and their applications. We place special emphasis on different design choices for key components of differentiable particle filters, including dynamic models, measurement models, proposal distributions, optimisation objectives, and differentiable resampling techniques.