A context-driven privacy enforcement system for autonomous media capture devices

The evolution of the Internet of Things and the related market has renewed the concept of media recording and sharing by means of a new kind of home-consumer devices, capable of continuous and autonomous media capture and upload. Such technologies have already begun to tamper with people’s privacy and discretion expectations, also raising many concerns about the potential legal implications. This work proposes an overall context-related privacy preserving system, based on context recognition. Our approach has been specifically developed considering contexts characterized by a high degree of similarity. The presented methodology has been devised in order to enforce privacy rules using image recognition techniques jointly with radio beacon technology. The reported results show that, in the peculiar environment considered in this paper, the use of radio beacon technology can help to increase the performances of image recognition techniques, both in terms of computational performances and elapsed time.