Dive Brief:
- AnyVision has partnered with NVIDIA to develop closed-circuit television (CCTV) surveillance cameras with facial recognition technology.
- The system will offer real-time analysis of streaming camera data, which AnyVision says can recognize a person in a crowd with up to 99% accuracy. Video information is sent to a system that uses algorithms to analyze people's faces and compare them with information in a central database.
- The partnership is part of NVIDIA's Metropolis Software Partner Program that launched in the fall, which aids software companies that offer production-ready surveillance solutions for smart cities.
Dive Insight:
This product has the potential to revolutionize law enforcement agencies' work, though it may also spook a lot of people and produce widespread cries of "Big Brother."
The technology continuously scans CCTV camera feeds 24/7, and it can work indoors and outdoors. It can be used on mobile devices as well as computers, which could allow police to, for example, check suspects against databases while in the field. The system could also further analyze suspects to find connections and clues based on their social media feeds.
AnyVision wants this innovation to help boost safety and security in cities. It can be used to track criminals or terrorists, or even to help find missing people. Consider how the facial recognition could help in instances such as the Boston Marathon bombing in 2013, when authorities captured surveillance video of the suspects whom they then worked to identify and track.
AnyVision claims that it will put high importance on protecting the personal data that the CCTV cameras capture. That will be important to create trust and get public buy-in for such a system. Cities looking to add facial recognition to existing surveillance systems might also consider establishing firm guidelines and regulations for using the technology and for captured information to alleviate public outcries about governments overstepping their bounds and becoming a nanny state.
