SPECIAL SESSIONS

In machine learning (ML) there is an insatiable demand for data. The success of modern ML algorithms, combined with the low cost of sensors, has led to experimental settings where observations are made jointly by multiple sensors that record different views or sense different modalities of the same scenario. Different views provide complementary information for describing and understanding a shared phenomenon. Multiview representation learning (MRL) is an emerging direction that builds on tools from machine learning and signal processing to distill meaningful features (or representations) from the joint information contained in multiple views. Machine learning and signal processing provide a wide variety of tools under the frameworks such as deep representation learning, coupled matrix and tensor decompositions, factor analysis and kernel methods for addressing those challenges.

The focus of this special session is novel approaches for multiview representation learning and their applications. Our aim is to bring researchers from both academia and industry working on complementary ideas in multiview learning under one umbrella to present their developments, and thereby encourage stimulating discussions about the current and future state of multiview research.

Organizers: Tanuj Hasija (Paderborn University, Germany), Tim Marrinan (Pacific Northwest National Laboratory, Seattle, WA, USA)

Bayesian methods are emerging as a powerful and flexible framework for many signal processing applications. The Bayesian framework provides principled ways to incorporate prior knowledge, perform inference and learning, and process uncertainties of any kind. Still, Bayesian methods have been challenging to scale up to high-dimensional data, making them computationally demanding for real-time applications.

This special session aims to bring together researchers and practitioners to present recent advances in efficient Bayesian methods for signal processing and their applications, including speech and audio processing, image processing, and biomedical image processing, among others. Moreover, the proposed special session will provide a platform for researchers and practitioners to foster collaborations on efficient Bayesian methods for signal processing.

Organizers: Bert de Vries (Eindhoven University of Technology, Netherlands), Francesco A.N. Palmieri, Giovanni Di Gennaro (Università degli Studi della Campania “Luigi Vanvitelli”, Aversa, Italy), Amedeo Buonanno (Italian National Agency for New Technologies, Energy and Sustainable Economic Development (ENEA), Portici, Italy)

The papers in each accepted Special Session will undergo the same review evaluation process of the regular papers submitted to MLSP 2023.

Special Session Proposals can be submitted via the Microsoft CMT Submission System. Please make sure to choose the appropriate session.

Please email your inquiry related to special sessions to: ieeemlsp-general@listserv.ieee.org.