Language switch: Deutsch
| Login
You are here:
Main content:

Dr.-Ing. Michael Gulbis

Michael Gulbis

Personal Data

Former PhD student (2005 - 2010)

Research Group Signal and Image Processing
Research: DFG-Postgraduate Program (GRK 466)

Research interests: Digital Audio Watermarking

Multimedia data has become a widely used carrier of information. One important difference to textual information is the greater degree of freedom with respect to the binary representation of the same information or meaning. While text data usually only allows one binary representation and changes therein immediately lead to errors or modifications of the perceived meaning, multimedia information may keep its meaning even after strong modifications of the binary file representing it. One example is lossy compression: The meaning of stored spoken language does not differ whether it is saved as PCM data or in a lossy compression format like mp3. Therefore, integrity protection of multimedia mostly refers to the preservation of the data content and not its binary representation.
In the area of multimedia integrity protection, digital watermarking has become an important technology. Depending on the required integrity level, several kinds of watermarking concepts are known: fragile, semi-fragile and content-fragile watermarks and further invertible watermarks. The recognition of digital audio data manipulation is a challenge addressed by various fragile and content-fragile watermarking algorithms. But so far none of the approaches provides satisfying results with respect to manipulation detection. Especially distinguishing malicious attacks from allowed post production operations is still an open issue.
My work is concentrated on the improvement of known methods and the development of novel techniques for content-based digital audio data authentication using digital watermarking.

Important areas of my research are:
  • audio data falsification detection and estimation
  • robust audio feature descriptions
  • high capacity audio watermarks