How Banks and Financial Institutions Use Face Recognition to Protect People, Property, and Assets
Stopping bank robbers might be the first thing that comes to mind when one thinks about physical security in the banking industry. But that’s only one thing that concerns banks and other financial institutions; they also have to protect their employees, branches, customers, and servers that contain customers’ personal data. Preventing and investigating incidents is easier now, thanks to sophisticated video surveillance networks that are used to conduct real-time monitoring of bank lobbies, ATM machines, vaults and other sensitive branch areas, as well as to gather evidence for post-incident investigations. While no one has enough time to precisely, fully and manually review the huge volumes of video footage created by those cameras, by implementing comprehensive video analytics software banks can leverage the power of Deep Learning and Artificial Intelligence to overcome these challenges, increase efficiency and drive effectivity.
The Advantages of Video Content Analytics
Video analytics software is able to identify, classify and index objects in surveillance video footage, thereby making it searchable, actionable and quantifiable. It detects, identifies, extracts and catalogs objects in video, and then drives analytic activities based on the extracted and classified data. To rapidly pinpoint and identify people of interest, video analysis includes filtered searches based on criteria such as gender, clothing type or color, accessories, vehicle type and color, direction of travel, and more. The analysis may be performed either in real-time, to enable customized alerts, or after the fact, to facilitate post-incident review of footage. These functions are highly effective in helping security teams accelerate post-incident investigations and increase real-time situational awareness. And, the technology is even more powerful when it includes “in the wild” facial recognition, a biometric technology that extracts faces detected in live or recorded video, then compares them to digital images in a watchlist.
Different Types of Facial Recognition Serve Different Purposes
Biometric technologies are becoming more widely used in the financial services industry to enhance public safety and physical security. For example, banks are increasingly using face recognition to authenticate a customer’s identity when s/he uses online banking applications. Banks also use “contained access control” facial recognition to manage entry and access for particular people to certain areas of a facility. Both of those applications involve having a known face image to match against only one other face in order to grant permission to a facility or a computer application. However, that technology alone is insufficient to track down unknown faces that may be involved in an incident: Security teams can leverage “in the wild” face recognition to enhance safety, prevent theft, and even provide information to other departments across the organization.
Accelerate Post-incident Investigations
First, let’s discuss how face recognition accelerates investigations. When a security breach or theft occurs, and investigators have a face captured on camera, a video content analytics system operator can search based on that image for face matches across footage from multiple cameras. This enables investigators to effectively search hours or days of footage in only a matter of minutes. One common problem for banks is identifying and tracking down card “skimmers.” Skimmers are criminals who install devices in ATM machines that capture cards’ magnetic stripe data and the associated PINs entered, then create counterfeit cards and use them to withdraw money from accounts. Banks usually have video cameras that record ATM activity, so when card skimming is reported or suspected, a bank can quickly review the footage during the relevant time period. Based on the video image of the potential skimmer, the bank can share that image with local law enforcement to help identify the card skimmer.
Real-Time Face Recognition Alerts for Suspicious Individuals
Now, let’s consider how real-time face recognition alerts can help resolve a criminal investigation, and prevent another crime. For example, a video analytics operator can set up alerting for when an unidentified individual is detected in several ATM cameras over a pre-defined, but short, period of time. This suspicious behavior might be indicative of an ATM skimmer at work. After reviewing the footage, the operator can determine whether a card skimming incident was likely to have taken place and add the suspect’s face image to the bank’s digital watchlist, to be alerted if face matches are detected in the future.
Similarly, system operators can set up a digital watchlist based on license plate recognition, if the ATM video camera is set up to capture images of vehicle license plates that pass by the ATM.
Preventing or Solving ATM Thefts
To prevent theft and other criminal activities, it is crucial to identify potential problem individuals when they enter the premises. For example, a bank may have a problem with thieves that linger inside or outside automatic teller machines (ATMs); perpetrators may wait for a legitimate customer to withdraw cash, then rob the customer. Video analytics software makes it possible to analyze footage of a theft incident, including faces of the suspect. Banks can then upload a digital image (either a still photo or video image) of the face to a watchlist on the video analytics platform. The next time a person with a matching face appears within a camera view inside or outside of that ATM — or another ATM in that video surveillance network — a central security team can be automatically notified in real-time, so they can quickly assess the match and determine how to respond. That person of interest could be proactively approached or monitored, and possibly apprehended if he or she is a potential security threat.
Branches and parking lots may have passcode-protected entrances; these are another security vulnerability because is possible for suspects to breach the premises in a car or on foot, by “tailgating” the car or person immediately ahead of them that entered via passcode. In this case, the bank could use either face recognition or license plate recognition to help identify the person of interest and set up a real-time alert for whenever that person or license plate number appears in a camera view.
Clearly, bank branches and corporate officers can save valuable time and reduce time to target by utilizing face recognition to alert on potential suspects, and to expedite the forensic review of post-incident investigations. But above and beyond traditional security uses, video analytics software and facial recognition technology can also benefit the customer service, planning, and operations departments, extracting valuable business intelligence from video footage, such as anonymous customer data, foot traffic, demographic data, and queue wait times.
Signup to receive a monthly blog digest.