The pairing of facial recognition software with artificial intelligence brings new meaning to the adage of being known by the company you keep.
The combination of two powerful technologies enables operators to pinpoint companions of a customer who draws attention as a suspected cheat or as a valuable player, says Henry Valentino, president and CEO of eConnect, which specializes in facial recognition software and data analytics.
The evolution of AI and facial recognition is a “game changer” for casinos or other sites with crowds of visitors, he said. “You’ll know everybody, because the technology is so good.”
To illustrate the new software’s capabilities, eConnect used video from surveillance cameras inside its Las Vegas headquarters. Technicians asked the software to show the people Valentino had been seen with at least seven times in the past 90 days. The software completed the task and produced two reports within a minute.
One display shows a hub-and-spoke design with Valentino’s picture in the middle and colored lines to photos of the 10 people who met the criteria. Each line specifies how many times Valentino met with that person, along with the person’s name if available; a line’s color indicates the level or type of association, such as frequency of contacts or how recently they occurred.
Taking the task a step farther, the software displayed photos of the people who had met at least seven times with each of Valentino’s 10 frequent contacts in the same time period. This more complex web provides direct and indirect connections, potentially useful for understanding wider social or operational networks, Valentino explained.
Searches can be based on various criteria, including:
- The time period to search, such as the past six months or the past 24 hours.
- The minimum number of meetings, to limit the listing of incidental contacts.
- The “degree of separation,” indicating family, friend, friend-of-a-friend, and so on.
- The classification of person being sought, such as employee or guest.
AI and facial recognition can flag not only bad actors and potential accomplices but also high-value customers, even those playing anonymously, and those who accompany them.
“eConnect’s been in the loss prevention world our entire existence. It started as a way to find the cheats and the groups operating in teams,” Valentino said. “But we’ve expanded.”
Even as casino gaming revenue continues to set records, he said, a large percentage of that comes from unidentified players, and “that’s where AI technology has come out of the surveillance room.” Players cite various reasons for gambling without their loyalty cards – they forgot it and don’t want to make the effort to get a new one, they left it in another machine, they simply want to stay anonymous.
“But with facial recognition and AI, essentially it doesn’t matter,” Valentino said. Another application for the technology is controlling employee access throughout the property, eliminating the need for individual key cards and increasing efficiency.
This year’s International Casino Surveillance Survey, cosponsored by the International Association of Certified Surveillance Professionals, found that 39 percent of properties use facial recognition technology. That reflects steady growth from the 25 percent reported in 2021, the first year of the survey. In addition, 64 percent of properties without facial recognition technology plan to add it, this year’s study found. Among properties using facial recognition, 28 percent said it had a “fairly large” or “huge” impact on operations. That’s double the rate of 2021. eConnect and Raving Consulting’s Tribal Gaming & Hospitality Magazine also cosponsor the annual survey.
“You still see safety and loss prevention, fraud prevention, and keeping crooks out as the primary uses for facial recognition,” Valentino said, but the ability to know everyone opens more possibilities.
Valentino said casino surveillance teams, often considered an expense, are reaching out to other departments about how a facial recognition system could pay off for them.
“Just like the internet,” he added, “eventually everybody’s going to get it.”
