I participated as a lecturer at the EU-COST funded PhD training school in De-identification for privacy protection in multimedia content, in Limassol Cyprus, in the beginning of October 2015.
De-identification for pricacy protection of multimedia content as speech, text, images, data is a new research area with the growing amount of available streaming data for research. The main research question is how to use data that contain individuals without revealing the identity of the person. One can blur faces in an movie but still the motion pattern of the person can be identified. One can also blur voices, but the way an individual expresses himself can reveal the person. There are several challenges around the privacy protection that has to be solved.
The training school had participants from the whole Europe and the USA, from the areas of image and movie processing, speech and text processing but also law.
EU will have several calls for research projects in this area next year