The amount of audiovisual (multimedia) content being produced, watched, edited, stored, broadcasted and exchanged in various application contexts is already phenomenal and quickly growing. At the same time, through all the exciting recent advances in digital signal compression, transmission and storage technology, multimedia has gained enormous potential in improving the traditional educational, professional, business, communication and entertainment processes.

In view of the above, there is a growing demand for tools capable of extracting the information from multimedia signals which describes the semantics of the content conveyed by these signals. Such tools should enable us, for instance, to easily access the events, persons, objects and scenes captured by the camera, to quickly retrieve our favourite themes from a large music archive or to efficiently generate overviews, summaries and abstracts of general multimedia documents. More insight in the needs for semantics-extraction tools can be found on our Application Page.

However, due to high data rates and rich content that is built of the information contained in different  modalities (pictures, sound and text), the problem of extracting semantics from multimedia signals is considerable.

The research efforts in the Multimedia Content Analysis cluster of the Information and Communication Theory Group address the above problem and aim at developing tools that will enable us to access the multimedia content semantics in a fast and reliable way, and in the applications belonging to the consumer, education, business and professional context.