- Facial recognition technology has advanced and can now be used to detect security events and enhance public safety
- VSBLTY’s state-of-the-art software provides facial recognition, weapon detection, and can also detect persons of interest, who represent an active threat
- VSBLTY is revolutionizing machine learning in retail digital displays to generate real-time analytics
A man parks his car near a shopping mall, goes to the trunk and takes out a loaded rifle. As he walks towards the building, he passes a camera mounted on a light post in the parking lot. The real-time video is run through a sophisticated algorithm that determines there is a 92 percent chance he is carrying a gun towards the mall. An alert is sent to the security manager, who pulls up the relevant video to look for himself and then hits an alarm that calls the police, closes and locks all external doors, and sends an alert to security personnel throughout the building. Police arrive and arrest the man before he fires his first shot.
“Our goal is to make people’s lives easier, and more secure.”
— Matt Pruitt, Chief Experience Officer, VSBLTY Groupe Technologies Corp.
In the same mall later that day, a young couple is shopping for ingredients to make their next best cocktail. A camera behind a digital display at the shelf runs their video through a similar algorithm, which picks up that they were holding a key bottle for the rather complex cocktail but are unsure and have put it back on the shelf. The system brings up a short video on the display, running through how to make the cocktail and the recipe. The couple buys the necessary bottles with the push of a single button and leaves the store happy.
On their way out of the store, the couple walks by a kiosk, where a digital ad for a soft drink pops up. The woman doesn’t like the ad, wrinkles her nose and turns away shaking her head. A camera behind the ad captures the motions and an algorithm registers that a woman in her 20s walked by the ad but didn’t like it.
Welcome to the intersection of marketing and security
In the examples above, all three scenarios used systems called Proactive Digital Display from Philadelphia-based VSBLTY, a leading software technology company, which has created sophisticated artificial intelligence and digital display technologies for retail and security applications in physical environments.
VSBLTY’s CEO, Jay Hutton, says the company’s retail and security software offers the same back-end technology but is distinct in their applications. The company’s security services typically use facial and object recognition, such as weapons identification algorithms, to deliver alerts to pre-designated personnel, who can then review footage and determine whether action needs to be taken.
The retail applications are a lot like what’s already happening when you shop online – but now VSBLTY is making this available in-stores.
“The reason the internet is winning the ad war is because those advertising impressions are measured,” Hutton explains. “I know, more or less, who sees my ad online. I know how long they spend on the page. I know what page they come from, what page they go to. All the breadcrumbs I can gather up is what makes this space so valuable to brands.”
However, advertisers and retailers haven’t been able to track data for in-store experiences the same way because the technology wasn’t there to do it – until now.
“Computer vision is adapting and evolving rapidly. It will be commonplace in the next couple of years,” Hutton adds.
Matt Pruitt has an extensive background in facial and pattern recognition from time spent with the FBI and conducting research into security for various divisions of the U.S. government. He joined VSBLTY in late 2019 as Chief Experience Officer, and is responsible for applying biometrics and video technology in the context of how it is experienced by people.
“I approach biometrics from a human experience perspective,” Pruitt says. “Our goal is to make people’s lives easier, and more secure.”
Formerly, he was involved in a project at several U.S. airports where customers can opt into using facial recognition instead of boarding passes to get on a plane – no paper pass or identification to bobble at the gate.
“It’s just easier to walk onto a plane,” he says. “Your face is your ticket. Biometrics, from a service standpoint, can really ease the pains consumers see. Critically, such systems are opt-in, involving only customers who have expressly signed up.”
He adds, “What’s exciting about the technology is how it can be adapted to specific use cases to meet the needs of people in a huge array of situations.” What that looks like is only starting to be explored.
“VSBLTY allows me to use biometrics technology in the public and get a sense of how people are interacting with it so we can adjust that experience,” Pruitt says. “Understanding how the public interacts with biometrics is very important.”
With privacy in mind, VSBLTY’s retail applications don’t use facial recognition or store any personal data. Rather, they use algorithms to analyze images in real-time to determine key demographics and mood data (male/female, age, happy/frustrated/mad) and then tailor the messages delivered.
It can be used in some pretty creative ways – and not just for mixing the latest cocktail.
VSBLTY recently collaborated with Tyson on the deployment of a new campaign featuring a variety of their products. They tested two sets of digital ads on digital freezer doors in grocery stores – all in separate markets. By measuring certain metrics such as how many people stopped to view the ad, touch the screen, sign-up for the store loyalty program, they quickly found which were the best performers. This allows brands to make smart, informed decisions about their national rollout strategy. It also informs the best performance for the greatest ROI on their digital display programs.
“This type use case, at the actual point of decision on purchase, is now possible to measure, in retail, for the first time,” Hutton says.
VSBLTY has worked on similar programs with cosmetic brands, detecting when people pick up an item and then playing videos about how to apply the make-up and offering to do a quick analysis of face type in order to recommend specific products.
“There is an educational component,” Hutton says. “The idea here is that education creates a lift in sales. In early tests, the metrics are off the charts. We are now at an inflection point, a movement north in this business.”
The emerging applications using such technology in a retail environment will have to overcome a perception of privacy issues if they are to be successful, Hutton adds.
“I think Facebook and Google were reckless and strident with the data entrusted to them. They got caught. We’re coming behind them,” he says. “In comes this next technology and it gets negatively coloured by the thing those guys are doing. I don’t think anyone in our industry is talking about this. We need to.”
“We are now at an inflection point, a movement north in this business.”— Jay Hutton, CEO, VSBLTY Groupe Technologies Corp.
Critically, personal data is only involved if law enforcement puts you on a list of people they’re looking for, or you opt into a retail application. However, in-store applications don’t store any images or videos and are completely anonymous and compliant with all privacy laws.
“Everything we do is real-time,” Hutton explains. “You are anonymous. I don’t know what you bought at 3:50 p.m. in the beverage store. What I do know is that a man in his 40’s bought Budweiser at that store. There’s more data on your phone that can be passively sent without your consent than I can pick up in retail with camera.”
Security applications for this digital AI technology are also being figured out in real-time – and with a lot of creativity based on individual situations. VSBLTY is working with several organizations such as schools, malls, and casinos to deploy weapon-recognition and facial-recognition systems to detect people about to commit a violent crime or who are wanted by law enforcement, but in some communities also finding a growing interest in ground-level, citizen-led programs.
How facial recognition plays a role in public safety in Mexico City
Mexico City is one of the most violent places on earth, Hutton says, with some 70 per cent of people having been the victims of violent crime in the last 10 years.
“It is a staggeringly tragic statistic,” he notes.
Working with city hall, VSBLTY is well into rolling out a program to place cameras in public areas across Mexico City to detect weapons and wanted criminals using facial and weapon recognition algorithms – but also at individual homes.
Mexico City residents interested in opting in request a subsidized package, which comes with a camera and the software needed to connect to the city-wide network and to view their own video feed in real-time on a computer at home. For privacy reasons, citizens cannot record video and don’t have access to the database of wanted criminals police have loaded into the system, but can report anything they are concerned about, which means they can contribute to the city-wide network of cameras watching for trouble and wanted individuals.
“So (in Mexico City) it’s bottom-up, all about citizen safety,” Hutton says. “People can opt into the program, and the government helps them pay for it. That’s refreshing.”
The home-based cameras have rapidly become a deterrent, with crime already declining 40 per cent in one neighbourhood with high rates of adoption.
Mexico City’s deployment is unique to the situation in that community and wouldn’t be appealing to, say, a Canadian suburb with lower crime rates.
In the future, such systems could be deployed to help find missing children or seniors with dementia – but only by authorities using approved images for that specific reason.
Pruitt says such biometrics technology has been around for more than 20 years, but it’s only in the last few years it has really become viable as a useful technology, as facial recognition technologies have moved from about 75 to 99.95 per cent accuracy.
“It’s really come a long way,” he says.
VSBLTY’s shares are listed on the Canadian Securities Exchange, and to learn more about the company and their advanced technology, visit their website here.