On March 4 2016, Sidsel Marie Nørholm of the Audio Analysis Lab will defend her Ph.D. thesis entitled “Enhancement of Speech – with a Focus on Voiced Speech Models”.

Thesis abstract: The thesis deals with speech  enhancement, i.e., noise reduction in speech signals. This has application in, e.g., hearing aids and teleconference systems. We consider a signal-driven approach to speech enhancement where a model of the speech is assumed and filters are generated based on this model. The basic model used in this thesis is the harmonic model which is a commonly used model for describing the voice d speech part of the speech signal. We show that it can be beneficial to extend the model to take inharmonicities or the non-stationarity of speech into account. Extending the model introduces extra parameters and we suggest methods to estimate these extra parameters and derive filters based on the extended models.

The Ph.D. assessment committee is

• Professor Stefani a Serafin (Chairman), Aalborg University
• Professor Ioannis (Yannis) Stylianou, University of Crete
• Reader Wenwu Wang, University of Surrey

The work was supervised by Jesper Rindom Jensen and Mads Græsbøll Christensen.

The defense will take place at 1 pm at Rendsburggade 14, room 3.429, 9000 Aalborg. The Department of Architecture, Design and Media Technology will host a reception after the defense.