Dissertation Jean-Marie Lemercier

1 March 2025, by David Mosteller

Photo: Lemercier

Jean-Marie Lemercier successfully defended his thesis "Model-based Techniques and Diffusion Models for Speech Dereverberation".

This thesis proposes different approaches to speech dereverberation by combining domain knowledge with modern machine learning. The work ranges from computationally lightweight predictive methods, through supervised generative diffusion models, and culminates in unsupervised diffusion models.

Latest articles

Photo: Fang

16.06.2025|SP

Dissertation Huajian Fang

Huajian Fang received his PhD degree for his thesis Model-Based Deep Speech Enhancement for Improved Interpretability and Robustness".

His thesis explores how integrating statistical modeling with deep learning-based approaches can improve speech enhancement. To achieve this, uncertainty estimation...

15.06.2025|SP

3 accepted Interspeech papers

We are pleased to announce the acceptance of the following papers for presentation at Interspeech 2025, which will be held in Rotterdam, the Netherlands, in August 2025:

Danilo de Oliveira, Julius Richter, Jean-Marie Lemercier, Simon Welker, Timo Gerkmann, "Non-intrusive Speech Quality Assessment...