The 16th ITG Conference on Speech Communication will take place from 24 to 26 September 2025 in Berlin with the theme"Speech Communication in the Age of Foundation Models". The conference is dedicated to current developments in theory, algorithms and applications of speech, audio and speech signal processing.
Our "Communications engineering" workgroup, headed by Prof Reinhold Häb-Umbach, is strongly represented at the conference. Presenting on 25 and 26 September:
- Alexander Werning, research assistant: "A fully Zero-shot Approach to Obtaining Specialised and Compact Audio Tagging Models"
- Adrian Meise, student of the group: "On the Application of Diffusion Models for Simultaneous Denoising and Dereverberation"
- Thilo von Neumann, research assistant: "Error Analysis in a Modular Meeting Transcript System"
Further information can be found here.
