Consolidating signal processing for the group room scenario in hybrid voice conferencing systems

• Applicants: Dr.-Ing. Aleksej Chinaev; Professor Dr.-Ing. Gerald Enzner
• Subject Area: Communication Technology and Networks, High-Frequency Technology and Photonic Systems, Signal Processing and Machine Learning for Information Technology
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 557407938

PC-based video conferencing proved to be surprisingly robust in terms of network, sound, and image quality during the years of restricted meetings and was discovered in this context as a provisionally viable replacement for face-to-face communication. The resource-saving added value of this technology is also of urgent importance for the future, but the inadequate support of hybrid operation consisting of some local participants on site and some remote participants is one of the biggest obstacles. Qualitative research provides coherent evidence of the urgent need for action to remedy notorious sound problems, such as acoustic feedback or level fluctuations in the group room. Concentrated microphone arrays only deliver limited relief. In our vision, participation with several conference devices in the group room should be a coherent user scenario in order to make digital value-added services (screen sharing, chat, minutes, etc.) available to the individual participants in the group room. This requires a fundamental analysis and provision of technologies for the electroacoustic linking of the conference devices in the group room to form a common voice into the conference. The project presented therefore makes suggestions for a fundamental engineering treatment using the means of digital audio signal processing. In doing so, we endeavor to be still minimally invasive to the infrastructure of a conference provider. In particular, two topological hypotheses for electroacoustic consolidation of the group room are pursued: a centralized processing paradigm in the connection server (i.e., with lossy transmission) and a distributed processing paradigm on the individual terminals in the group room. The challenging questions for digital signal processing in each case are the fine sampling rate synchronization of the otherwise autonomous devices in the group room and the multi-channel acoustic consolidation of the relevant microphone signals with various room transfer functions and potentially background noise. In the proposed project, these questions will be focused on two applicants (one of whom is a first-time applicant for own position) and carried out cooperatively at the same location. The work program of the project describes the different as well as the cooperative tasks in the context of the mentioned conference topologies with and without the use of a connection server. However, the proposed project has fundamentally different requirements, on the one hand due to the lossy low-rate transmission to a connection server and on the other hand due to the distributed multi-channel signal processing in the group room.

• DFG Programme: Research Grants