ARI GUEST TALK: 22.5.2024 3pm
Date: Wednesday, 22nd May 2024, 3pm.
Venue: Austrian Academy of Sciences, Georg Coch-Platz 2, 3rd floor, SE-room 1, 1010 Vienna.
In his 1938 dissertation, Herman Wold developed a systematic decomposition of stationary stochastic processes into a singular component (a sum of sinusoids) and a regular component (white noise passed through a stable and causal linear filter). This decomposition has been the implicit foundation of many models developed for speech signals such as sinusoidal models, linear predictive models, and harmonic+noise models. In this talk, we will discuss how this modelling paradigm can be used as a systematic foundation for random wave fields and for unravelling the mysteries of diffuse and modal sound field components.
Gernot Kubin is Founding Director of the Signal Processing and Speech Communication Laboratory at TU Graz, Austria, since 2000, and holds a Dr.techn. degree (sub auspiciis praesidentis) from TU Vienna. Earlier appointments include: CERN Geneva, TU Vienna, Philips Research Labs Eindhoven, AT&T Bell Labs Murray Hill, KTH Stockholm, Vienna Telecom Research Centre FTW, Global IP Sound Stockholm and San Francisco, UC San Diego, Danang UT, and TU Munich. His research interests are in nonlinear signal processing as well as speech and audio communication. He was General Chair for INTERSPEECH 2019 and is a member of the ISCA Advisory Board since 2020.