(Image Credit: iStockPhoto/alice-photo)
A new initiative from the US Office of the Director of National Intelligence wants to analyse video feeds in real-time in order to spot behaviour which could be deemed as suspicious. The project is called Deep Intermodal Video Analytics (DIVA) and it will be a joint effort between academics, the private sector, and ODNI’s Intelligence Advanced Research Projects Agency.
You don’t want to be flagged every time you give a piece of gum to a friend
Senior figures in surveillance agencies have brought to light the dangers of ‘data deluge’ whereby too much information and not enough manpower can make it difficult to find credible threats, but if the data can be paired with a system which monitors suspicious behaviour then it could be helpful.
“The DIVA program will produce a common framework and software prototype for activity detection, person/object detection and recognition across a multi-camera network,” IARPA officials explain. “The impact will be the development of tools for forensic analysis, as well as real-time alerting for user-defined threat scenarios.”
Few will argue that surveillance measures aren’t required to keep us safe, but to what extent we should give up our freedom and privacy is a matter of constant debate. Such a system will be seen as intrusive, and will create an outcry from privacy activists.
Too much information and not enough manpower can make it difficult to find credible threats
DIVA would have to be robust but not add to current information overload problems. The system will be looking for obvious threats – like someone carrying a firearm – but also for people exchanging objects or leaving a suspicious package somewhere. There are bound to be some brilliant minds working on the algorithm behind DIVA, but you don’t want to be flagged every time you give a piece of gum to a friend or take out your trash.
As famously said by French 17th-century clergyman and politician Cardinal Richelieu: “If one would give me six lines written by the hand of the most honest man, I would find something in them to have him hanged.”
Part of the algorithm will be detecting suspicious behaviour prior to an event taking place, but with everyone so different in their mannerisms, there is sure to be some degree of error. To begin with, the researchers will focus on standard indoor and outdoor CCTV footage. If successful, DIVA will be extended to mobile cameras.
What are your thoughts on mass behaviour monitoring? Let us know in the comments.