Video has become the single most powerful medium in the world of communications. Unlike any other data source, it has the power to give us real-time information about human behavior and how people interact with the world and transform our ability in a highly dynamic environment to proactively protect our citizens.
The private sector is on the leading edge of leveraging rapidly evolving disruptive technology to refine their video marketing campaigns, improve their demographic targeting, and extract increasing value from video. Governments at all levels, and national security and law enforcement agencies in particular, need to catch up to these transformative technologies as they tackle three critical challenges in the face of the massive explosion of video intelligence (VIDINT):
- Assessing in real time exactly who and what is occurring in the massive amounts of video collected in the local environment, border crossings, and via intelligence operations
- Locating and identifying video material disseminated by individuals with nefarious intent, both openly and in the deep web, to better understand their audiences and counter or locate them
- Developing tactical and strategic information operations leveraging video and mechanisms to track their impact
Existing facial recognition, machine learning and artificial intelligence capabilities in federal, state and local government are no longer sufficient and require a disproportionate amount of human resources with a relatively high degree of false positives and potential for human error. In a world of increased threats to our safety and security coupled with limited human and fiscal resources, you cannot afford to waste time manually tracking and analyzing this data can be better directed to other critical missions. These search technologies looks at the metadata stored with videos while other technologies leverage image techniques (like classification) to find objects, cars, or company logos in videos.
True video analytics involve using computer vision algorithms to analyze video pixels automatically, and over time — not only identifying objects in the scene, but also tracking their movements and behavior. With true video analytics you can track patterns of movement of hundreds of objects in the scene, as well as their size, shape, speed, and direction of movement.
The Intelligent Video-fingerprinting Platform (IVP) provides true video analytics powered by computer vision that will enable national security and law enforcement agencies to drive operations based on real-time analysis of video rapidly integrated with other big data sources.