Our research activities on signal processing and transmission focus on the Our research activities on signal processing and transmission focus on the development of new algorithms for efficient compression of image and video signals. Special attention is given to coding of video signals at ultra HD resolution with high fidelity as well as to scalable and error robust video coding. We also work on quality improvement of noisy and erroneous video data and we develop methods for transcoding of video data between different compression standards. In the field of image and video analysis we work on extraction of reliable features for detection and tracking of objects. Signal processing for multi-view camera systems forms another interesting research topic where we develop algorithms for scalable view representation and archiving as well as for view interpolation at the receiver.
Completed Research Topics
Methods for optimized video and multimedia transmission over error prone channels
Transcoding of Digital Videostreams
Digital Watermarking of Multimedia Data
Video coding taking into account the human visual system
Scalable Vidoe Streaming for Wireless Video-On-Demand Services
Models and methods towards compressed video analysis and interpretation for adaptive presentation provisioning
Quality enhancement of mobile recieved video data
Special Coding techniques for Computer Animated Videos
Scalable Multi-View Video Coding
Mobile In-Car TV-Reception
View-independent automatic flight guidance
In the audio group we focus on theoretical development as well as experimental verification of algorithms for acquisition, enhancement, analysis, terpretation, and rendering of audio and speech signals. Various longstanding allenges, such as acoustic scene analysis, natural barrier-free human/machine terfaces, e.g., for telecommunications and immersive multimedia, or telligent binaural hearing aids and perfect synthesis of sounds and vefields, form the background for research on highly advanced signal ocessing methods. Characteristically, the exploitation of the spatial mension by multichannel signal processing algorithms plays a key role in many our projects.
Die physikalische Modellierung zeit- und ortsabhängiger Probleme mit den Mitteln der Systemtheorie führt auf diskrete Modelle, die sich effizient am Rechner implementieren lassen. Aktuelle Anwendungen aus dem Multimediabereich sind Verfahren zur digitalen Klangsynthese und zur Berechnung und Visualisierung von akustischen Schwingungen. Die Analyse zeit- und ortsabhängiger Signale eröffnet auch Wege zur Objektlokalisierung in audiovisuellen Szenen.