Invited Talk
Prof. Dr.-Ing. Nilesh Madhu (IDLab / Ghent University)
“All the better to understand you with”
The exciting opportunities and open challenges in listening augmentation
Effective communication is important for a harmonious, well-adapted society. This requires a clear and timely exchange of messages and intent. Especially in the field of tele-communications, engineers have been working on this problem for decades and, given the wide range of challenges in this field, they will remain busy for several years to come. In this talk I shall introduce the 3 fundamental hurdles to an effective communication and show how almost all of speech- and audio-related research is focused on tackling these hurdles. Further, I shall present how the solutions to these challenges fit a standard machine learning framework and demonstrate these solutions in practical scenarios encountered in daily life. Lastly, I will briefly reflect upon the value of domain knowledge and simplified stochastic models, and their role in the data-driven algorithmic landscape of today. The ultimate goal of the talk is not only to develop a good understanding of the open challenges in the field of speech communication, but also to stimulate discussion on how to best leverage current progress in deep learning to overcome these challenges.
Nilesh Madhu is professor for audio, speech and signal processing at Ghent University, Belgium. He is passionate about signal processing and is especially interested in its applications in the fields of communications, healthcare, and automation. A key focus of his group is the exploitation of domain knowledge within data-driven, deep-learning based approaches, yielding explainable algorithms for robust signal detection and enhancement.
He was granted his Dr.-Ing. degree (summa cum laude) from the Ruhr-Universität Bochum in 2009. His dissertation was on algorithms for the localisation and separation of acoustic sources using microphone arrays. Following this he was awarded a Marie-Curie fellowship for a two-year postdoctoral stay at the KU Leuven, Belgium, where he gained expertise in the fields of hearing prostheses and biomedical signal analysis. During his industry tenure at NXP Semiconductors, Belgium, he held the position of principal scientist within the product line Mobile Audio Solutions. He and his team successfully developed beyond state-of-the-art algorithms for audio and speech enhancement, which are incorporated in mobile devices of major OEMs today.
