Video Annotation of TV Content using Audiovisual Information

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


Content provider companies and televisions have largevideo archives but usually do not take full advantage of them. Inorder to assist the exploration of video archives, we developedViewProcFlow, a tool to automatically extract metadata fromvideo. This metadata is obtained by applying several state of theart video processing methods applied to a real world challenge:cut detection, face detection and object identification. In addition,we also developed a method to annotate videos with conceptsfrom audio and visual information. The main novelty of thistechnique is the use of environmental sound recognition toannotate video. The goal is to supply the system with moreinformation such that it has a better understanding of the contentand also to enable better browse and search functionalities. Thistool has been included in the workflow of a video productioncompany, which confirms its success.
Original languageUnknown
Title of host publicationProceedings of the IEEE International Conference on Multimedia Computing and Systems
Publication statusPublished - 1 Jan 2012
EventIEEE International Conference on Multimedia Computing and Systems -
Duration: 1 Jan 2012 → …


ConferenceIEEE International Conference on Multimedia Computing and Systems
Period1/01/12 → …

Cite this