Search

Talk Invitation: Bayesian Estimation Using Learned Models: Application to Multi-Object Tracking

Telecommunications / News / Talk Invitation: Bayesian Estimation Using Learned Models: Application to Multi-Object Tracking
13. August 2026
Guest speaker Florian Meyer from University of California San Diego

Speaker:

Florian Meyer
University of California San Diego (Scripps Institution of Oceanography & Department of Electrical and Computer Engineering), La Jolla, CA, USA

Date & Time:
Thursday, August 13, 2026 at 11:00

Location:
Lecture Hall EI 4 (Reithoffer-Hörsaal), TU Wien, Gußhausstraße 25, 1040 Vienna

Abstract:

Bayesian estimation methods are central to applications including oceanography, autonomous navigation, radar and sonar sensing, and undersea surveillance. By combining physical models with probabilistic inference, they offer important advantages over end-to-end data-driven methods, including interpretability and principled uncertainty quantification. Their performance, however, can deteriorate when simplifying assumptions in the statistical model fail to capture the true data-generating process—for example, because of unmodeled environmental or propagation effects.

This talk presents neural-enhanced Bayesian methods for multiobject tracking in complex sensing environments. Rather than replacing Bayesian inference with an end-to-end neural architecture, the proposed approach introduces learned components only in parts of the statistical model that are likely to be inaccurate, while retaining the structure and uncertainty quantification of the underlying Bayesian formulation.

We first discuss supervised neural enhancements to detect-then-track methods and then focus on a track-before-detect framework that operates directly on raw sensor data. In the latter framework, a neural-network-based normalizing flow parameterizes a probability density function for the received signal intensity associated with each object, thereby capturing unmodeled propagation effects. Factor-graph inference estimates object states and existence, while an expectation–maximization procedure updates the normalizing-flow parameters online using the resulting posterior distributions, without requiring labeled data or a separate training dataset.

Results using simulated and real radar and sonar data demonstrate improved multiobject localization and tracking and indicate that the learned probability model reflects physically meaningful characteristics of the propagation environment.

About the speaker:

Florian Meyer received the M.Sc. and Ph.D. degrees (with highest honors) in electrical engineering from TU Wien, Vienna, Austria, in 2011 and 2015, respectively. He is an Associate Professor with the University of California San Diego, La Jolla, CA, USA, jointly between Scripps Institution of Oceanography and the Electrical and Computer Engineering Department. From 2017 to 2019, Dr. Meyer was a Postdoctoral Fellow and Associate with the Laboratory for Information & Decision Systems, Massachusetts Institute of Technology, Cambridge, MA, USA, and from 2016 to 2017, he was a Research Scientist with the NATO Centre for Maritime Research and Experimentation, La Spezia, Italy.

His research interests include statistical signal processing, high-dimensional and nonlinear estimation, inference on graphs, and estimation using learned models.

He is the recipient of the 2021 ISIF Young Investigator Award, a 2022 NSF CAREER Award, a 2022 DARPA Young Faculty Award, and a 2023 ONR Young Investigator Award. He is currently an Associate Editor with IEEE Transactions on Signal Processing.

About Cookies

Our website uses cookies to ensure you get the best experience on our website, for analytical purposes, to provide social media features, and for targeted advertising. This it is necessary in order to pass information on to respective service providers. If you would like additional information about cookies on this website, please see our Data Protection Declaration.

  • These cookies are required to help our website run smoothly.

    Name Purpose Lifetime Type Provider
    wordpress_test_cookie Testing-Cookie to check whether cookies are allowed. 1 Year HTTP Homepage TUW
    PHPSESSID Used by WordPress to retain the state of your current user session for all page requests. Session HTTP Homepage TUW
    wordpress_logged_in_{hash} Used by Wordpress to keep users logged in. {hash} represents an unique user token. 1 Year HTTP Homepage TUW
    wp-settings-time-{id} Used to customize your view of admin interface, and possibly also the main site interface. 1 Year HTTP Homepage TUW
    wordpress_sec_{hash} This cookie is used to store your authentication details. Its use is limited to the admin console area. {hash} represents an unique user token. 1 Year HTTP Homepage TUW
    wp-settings-{id} Used to customize your view of admin interface, and possibly also the main site interface. 1 Year HTTP Homepage TUW
    wp-wpml_current_language Stores the current language. This cookie is enabled by default on sites that use the Language filtering for AJAX operations feature. 1 Day HTTP Homepage TUW
    wp-wpml_current_admin_language_{hash} Stores the current WordPress administration area language. {hash} represents an unique user token. 1 Day HTTP Homepage TUW
    CookieConsent_117a3e Saves your settings for the use of cookies on this website. 1 Year HTML Homepage TUW

  • These cookies help us to continuously improve our services and adapt our website to your needs. We statistically evaluate the pseudonymized data collected from our website.

    Name Purpose Lifetime Type Provider
    _pk_id.136.56ce Used to store a few details about the user such as the unique visitor ID. 13 months HTML Matomo TUW
    _pk_ref Is used to store the information of the users home website. 6 months HTML Matomo TUW
    _pk_ses.136.56ce Is needed to store temporary data of the visit. 30 minutes HTML Matomo TUW