The Multimedia Signal Processing and Security Lab has a strong emphasis on development of state of the art techniques in visual data processing applicable in a real world context, often in a multi-disciplinary setting. For example, we have published on visual techniques in fish aquaculture, in wood industry (i.e. wood log tracking), automated material hardness measurements, and Blu Ray copy protection systems. Among the application areas considered, the analysis of bio-medical signals has been one of the main topics of interests in the last years: brain-MRI analysis for epilepsy diagnosis, colonoscopy analysis for staging of colon polyps, analysis of duodenal endoscopy data for diagnosis of celiac disease, analysis of various signal types for biometric recognition (e.g. vascular biometrics, iris recognition, fingerprint recognition, EEG-based recognition).
Due to the recent progress in machine learning and artificial intelligence, we are very interested in the relation between those new approaches and the traditional model-based ones, and to utilize the respective communalities and differences in the development of new, eventually hybrid techniques. Also, the question for the generation of appropriate domain-specific training data, while respecting the privacy demands and regulations, is of particular interest for our research.
