Viewdle has announced a partnership with Reuters for facial recognition video index and search. The company showcased their Viewdle Engine today TechCrunch 40, check it out here.
According to an e-mailed release “Reuters.com, will use the Viewdle engine to automatically extract metadata from select Reuters satellite streams and provide real-time indexing and search capabilities.”
I first heard about this technology several months ago at the last OMMA Video event. It fills a critical need for companies like Reuters which constantly monitors and serves online video featuring public figures.
Much of the workflow for professional online video has been digitized but adding metadata during ingestion is one area of continued weakness. Individual loggers are still heavily depended on to classify content, a time-consuming, manual process. In a newsroom improved speed and accuracy as a result of this type of technology offers a big advantage.
The beauty of Viewdle is that once a person has been recognized, he or she can be located in all past and future content automatically. A producer putting together a package on say, George W. Bush, can find every piece of George Bush content in the Reuters Library instantly, and can do so for new content much more quickly.
Where this gets really exciting is integrating it with other search technologies to form a completely automated digital indexing system. Manual ingestion and logging will soon be a thing of the past, replaced by an entirely automated digital workflow in which content is instantly available to producers and editors as soon as it is shot.