-
Dynamically Iterated Filters: A unified framework for improved iterated filtering and smoothing
Authors:
Anton Kullberg,
Martin A. Skoglund,
Isaac Skog,
Gustaf Hendeby
Abstract:
Typical iterated filters, such as the iterated extended Kalman filter (IEKF), iterated unscented Kalman filter (IUKF), and iterated posterior linearization filter (IPLF), have been developed to improve the linearization point (or density) of the likelihood linearization in the well-known extended Kalman filter (EKF) and unscented Kalman filter (UKF). A shortcoming of typical iterated filters is th…
▽ More
Typical iterated filters, such as the iterated extended Kalman filter (IEKF), iterated unscented Kalman filter (IUKF), and iterated posterior linearization filter (IPLF), have been developed to improve the linearization point (or density) of the likelihood linearization in the well-known extended Kalman filter (EKF) and unscented Kalman filter (UKF). A shortcoming of typical iterated filters is that they do not treat the linearization of the transition model of the system. To remedy this shortcoming, we introduce dynamically iterated filters (DIFs), a unified framework for iterated linearization-based nonlinear filters that deals with nonlinearities in both the transition model and the likelihood, thereby constituting a generalization of the aforementioned iterated filters. We further establish a relationship between the general DIF and the approximate iterated Rauch-Tung-Striebel smoother. This relationship allows for a Gauss-Newton interpretation, which in turn enables explicit step-size correction, leading to damped versions of the DIFs. The developed algorithms, both damped and non-damped, are numerically demonstrated in three examples, showing superior mean-squared error as well as improved parameter tuning robustness as compared to the analogous standard iterated filters.
△ Less
Submitted 17 April, 2024;
originally announced April 2024.
-
On the Relationship Between Iterated Statistical Linearization and Quasi-Newton Methods
Authors:
Anton Kullberg,
Martin A. Skoglund,
Isaac Skog,
Gustaf Hendeby
Abstract:
This letter investigates relationships between iterated filtering algorithms based on statistical linearization, such as the iterated unscented Kalman filter (IUKF), and filtering algorithms based on quasi-Newton (QN) methods, such as the QN iterated extended Kalman filter (QN-IEKF). Firstly, it is shown that the IUKF and the iterated posterior linearization filter (IPLF) can be viewed as QN algor…
▽ More
This letter investigates relationships between iterated filtering algorithms based on statistical linearization, such as the iterated unscented Kalman filter (IUKF), and filtering algorithms based on quasi-Newton (QN) methods, such as the QN iterated extended Kalman filter (QN-IEKF). Firstly, it is shown that the IUKF and the iterated posterior linearization filter (IPLF) can be viewed as QN algorithms, by finding a Hessian correction in the QN-IEKF such that the IPLF iterate updates are identical to that of the QN-IEKF. Secondly, it is shown that the IPLF/IUKF update can be rewritten such that it is approximately identical to the QN-IEKF, albeit for an additional correction term. This enables a richer understanding of the properties of iterated filtering algorithms based on statistical linearization.
△ Less
Submitted 20 November, 2023; v1 submitted 14 September, 2023;
originally announced September 2023.
-
Predicting EEG Responses to Attended Speech via Deep Neural Networks for Speech
Authors:
Emina Alickovic,
Tobias Dorszewski,
Thomas U. Christiansen,
Kasper Eskelund,
Leonardo Gizzi,
Martin A. Skoglund,
Dorothea Wendt
Abstract:
Attending to the speech stream of interest in multi-talker environments can be a challenging task, particularly for listeners with hearing impairment. Research suggests that neural responses assessed with electroencephalography (EEG) are modulated by listener`s auditory attention, revealing selective neural tracking (NT) of the attended speech. NT methods mostly rely on hand-engineered acoustic an…
▽ More
Attending to the speech stream of interest in multi-talker environments can be a challenging task, particularly for listeners with hearing impairment. Research suggests that neural responses assessed with electroencephalography (EEG) are modulated by listener`s auditory attention, revealing selective neural tracking (NT) of the attended speech. NT methods mostly rely on hand-engineered acoustic and linguistic speech features to predict the neural response. Only recently, deep neural network (DNN) models without specific linguistic information have been used to extract speech features for NT, demonstrating that speech features in hierarchical DNN layers can predict neural responses throughout the auditory pathway. In this study, we go one step further to investigate the suitability of similar DNN models for speech to predict neural responses to competing speech observed in EEG. We recorded EEG data using a 64-channel acquisition system from 17 listeners with normal hearing instructed to attend to one of two competing talkers. Our data revealed that EEG responses are significantly better predicted by DNN-extracted speech features than by hand-engineered acoustic features. Furthermore, analysis of hierarchical DNN layers showed that early layers yielded the highest predictions. Moreover, we found a significant increase in auditory attention classification accuracies with the use of DNN-extracted speech features over the use of hand-engineered acoustic features. These findings open a new avenue for development of new NT measures to evaluate and further advance hearing technology.
△ Less
Submitted 27 February, 2023;
originally announced February 2023.
