Wednesday, 21 Feb 2018
You are here:

Subscribe to our Newsletter

How many eyes has a typical person? (ex: 1)

Twitter Feed

EMail Print


Virginia Fernández Arguedas & Krishna Chandramouli (QMUL/VIT) visiting Prof. Atta Badii (UoR)
23.11.2012 - 23.11.2012
Internal Visits


In this visit, Virginia Fernández Arguedas presented a system to populate multimedia ontologies with an object tracker and an object classifier, providing the ontology with enough information to infer and reason basic events, such as people walking together. As a result, a connection between video analytics systems and a sophisticated surveillance knowledge representation was presented. Smart surveillance systems require powerful video analytics to detect, index and retrieve information for real-time surveillance as well as for forensic application. However, the VideoSense project goes further beyond, considering the implications of ethical, legal and privacy implications within surveillance systems. During Dr. Krishna Chandramouli’s visit from the 6 to the 24 November 2012 , a unified surveillance knowledge representation was designed.

The proposed knowledge representation defined a unified structured modulated into several components, such as multimedia, geolocation geometrical view, object, event, neighbourhood analysis and time ontologies. QMUL participated in the design and played a relevant role on its development for surveillance applications. During this visit, QMUL and UoR discussed the opportunities for VideoSense.

Dr. Krishna Chandramouli and Virginia Fernández Arguedas presented the possibility of extending the ontology to multiple surveillance scenarios proposing the collaboration of all members from the VideoSense Consortium, and specifically of UoR, for the ontology population and use case design. Moreover, QMUL and UoR discussed QMUL’s role and collaboration towards the Context Aware Privacy Mini-project led by UoR, exploiting QMUL work on object classification for surveillance applications.