Workshop 'Meet the PhD Jury: Advances in Acoustic Signal Processing'

We are delighted to announce a keynote by Professor Emanuël Habets (jointly affiliated with Fraunhofer IIS and the Friedrich-Alexander Universität Erlangen-Nürnberg) on the topic of "Neural Directional Filtering and Dynamic Slimmable Networks".

Neural Directional Filtering is a novel approach to far-field speech capture that utilizes deep neural networks for precise directivity control. The approach outperforms traditional linear and parametric filtering approaches, enabling high spatial selectivity with a small number of microphones. Dynamic Slimmable Networks address the computational inefficiency of conventional separation networks. By adjusting computational effort to the input characteristics, significant complexity reduction is obtained, while maintaining separation performance.

This will be followed by about 6 short presentations, delivered by international researchers from Ghent University, Aalborg University and the University of Hamburg.