FB pixel

Facial recognition and video analytics in CCTV meet with mixed feedback

Nice enables collaboration between UK businesses and police
Facial recognition and video analytics in CCTV meet with mixed feedback

Israel-based enterprise software solutions provider Nice has recently partnered with the UK’s National Business Crime Centre (NBCC) to assist businesses and police with investigations using its biometric Digital Evidence Management Software (DEMS).

The move potentially paves the way for an improved way of intelligence-gathering and more efficient policing through CCTV video feeds and other evidence shared with law enforcement by retailers and other businesses, but also raises fresh privacy issues related to the collection of biometrics and other information by private organizations.

Participating organizations can register their CCTV feeds with the NBCC through Nice Investigate’s Public Portal, allowing law enforcement officers to review footage, and potentially run forensic facial recognition queries.

DEMS can process not only face information from CCTV footage, but also other forms of digital content, which can then be sifted through and analyzed using Microsoft Azure.

A month after the beginning of the new partnership between Nice and the NBCC on processing data collected by private businesses, we analyze its potential and reflect coverage and opinions from the general public about it.

Using face recognition to improve business safety

According to ITPro, pharmacy chain Boots was one of the first large companies to share its CCTV data with DEMS, with the company believing the move will help them “get better at reporting crimes.”

Also talking to the technology publication, Tony Porter, Corsight AI’s CPO and former British Surveillance Camera Commissioner echoed the point, explaining that he believes the UK public would support the broader use of the system if the appropriate protections are put in place.

Recent data from NW Security Group suggests the adoption of advanced video analytics will continue to grow in the next few years. At the top of this list is facial recognition, behavior or event-based analytics, ANPR, video motion detection (VMD), object tracking, object detection and classification, directional detection, and OCR (optical character recognition).

Addressing privacy concerns

A separate study by Enterprise Times shows similar adoption rates of advanced video analytics, also highlighting how the pandemic has brought about a new series of crowd tracking and prediction systems.

The same survey, however, suggests one of the challenges for surveillance system operators remains to balance privacy with protecting business and the public.

For instance, the data showed that the video shaping feature used by Amazon Ring cameras still captured the whole field of view, with unfiltered images potentially still being extracted.

Collection and storage of data captured by surveillance is also a recurrent issue, the survey suggested, with very little signage in large cities having adequate and up-to-date information on it.

The industry’s view

A recent post on IPVM (subscription required and recommended), a video surveillance industry website, has raised issues connected to identifying known thieves in-store using facial recognition.

“Why shouldn’t they be allowed to use facial recognition to identify a known thief in their store?” asked one reader in kicking off the conversation. “So you’re also giving a thief the right to privacy so they can rob you?”

IPVM founder John Honovich replied to the post by asking additional questions related to what retailers should do if they get a match.

“Can they be certain the match is correct? And if you are certain, what are you going to do? Having the store manager apprehend a criminal? Call the police and hope they come in time?”

Additionally, Honovich explained that while London Police have long grappled with the risks of apprehending suspects, facial recognition makes this even more difficult.

“Some proponents of facial recognition in retailers had advocated detecting all sorts of criminals in stores (e.g., a child molester walks into Walmart and an alert is generated),” Honovich explained.

“For [Organized Retail Crime] or shoplifting when the person may actually commit a crime then and there, I see the potential. For general criminal alerts, this strikes me as high risk (‘Excuse me sir but, no offense, my computer system says you look like a child molester’),” he added.

While there is no consensus over these issues, it is interesting to look at the debate, and how it continues to evolve as facial recognition solutions become more and more accurate.

What are your thoughts about these issues? Let us know in the comments below.

Related Posts

Article Topics

 |   |   |   |   |   |   |   |   |   |   | 

Latest Biometrics News

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Most Read This Week

Featured Company

Biometrics Research

Biometrics White Papers

Biometrics Events

Explaining Biometrics