Of the many technologies that have transformed the world we live in, cloud computing decidedly led the digital transformation – and its evolution is far from over. Ján Lunter, founder and CEO at Innovatrics, sheds light on cloud-based biometrics and how they can help streamline identity management.
With an annual growth rate (CAGR) of 17.5 percent, the cloud computing market is expected to reach a staggering $832 billion by 2025.Opens a new window The cloud allows companies to outsource their IT services and store millions of data points securely online. With pay-as-you-go services, companies can reduce the costs of expensive hardware acquisition or software maintenance.
Additionally, storage capacity has been skyrocketing – granting new use cases for numerous business areas, including identity management. A few years ago, full-scale biometric identification was an incredibly challenging task for the cloud. The data, especially for real-time video streaming, was simply too large. Today, edge computing and specialized systems on chip (SoC) disrupt identity management.Â
Let’s dive into the prospects of cloud computing in the identity management market and how it advances both public and private security.
A Fresh Breeze for Identity ManagementÂ
Biometric technology – the use of characteristics like iris, fingerprints, or faces to validate people’s identity – requires the analysis of millions of data points. These vast amounts of data need to be stored safely to give access to people who need access to a service or location without endangering their physical or information security.Â
Whereas cloud computing has been around for some time, edge computing facilitates time-sensitive data processing even without internet access or with limited network bandwidth. By locating devices (such as observation cameras or smartphones) close to the central processing systems, edge computing permits fast and streamlined operations.
Specialized systems on chips – that are small but computationally strong – can process video data before sending them to the cloud for final distillation. Biometric security systems usually need a robust server to stream video in real-time, with several GB of RAM required for each stream. With specific SoCs like Ambarella or edge devices such as nVidia Jetson or Blaize, the computing power can be moved directly to a standard surveillance camera (or near to it). Through on-camera processing, the biometric algorithm can detect faces and convert them into so-called templates – much smaller files than photos themselves. These templates are then sent to the cloud and used for comparison purposes. The necessary bandwidth for authenticating a user drops by orders of magnitude as the final matching process happens in the cloud or on a lower-end console.Â
New Use Cases for Biometric Identification
These enabling technologies open the biometric identification and verification market to the popular Software-as-a-Service (SaaS) model. By plugging cameras into edge computers (or buying smart cameras with SoCs integrated), companies can leave it to the installed system to do the rest. Ultimately, this reduces hardware requirements and maintenance costs significantly.
SoC and edge processing in facial recognition will simplify access control and enhance public security. The technologies enable new, innovative deployment options by significantly cutting down network bandwidth and resources at the central site.
For example, public authorities can keep premises secure by monitoring trespassers. The cameras can control gates in gyms or within gated communities to make sure just authorized people enter. Facial recognition algorithms can detect faces, including those covered with facemasks or blurred due to night vision. New behavioral algorithms like human appearance detection and tracking add another layer of security – they can identify suspiciously behaving people and, for example, send a notification to the police squad near the site. Only the police can then identify the trespasser if they are indeed trespassing.
Smart glasses can now run biometric identification, so a security guard can wear them in crowded places and use the identification in real-time. The glasses detect all faces, extract templates and send them to the cloud via wi-fi. For example, anyone who is not a registered employee or worker in an office building is immediately tagged. The guard then can help them with finding the correct person or office.
See More: Using Biometrics to Create Personalized Yet Secure Customer Experiences
Is Sensitive Data Safe on the Cloud?Â
One rising (and justified) concern of users is data protection. The way data is secured is of utmost importance when using digital storage. While cloud data is usually highly encrypted and can be secured with reliable cybersecurity measures, deploying additional measures is also necessary.
The big advantage of using edge computing is that templates are the only data traveling to the cloud. The templates represent the picture of a person, but they cannot be reverse-engineered to produce the source photo. So, even if hackers incept the photos, personal data won’t be endangered.
Having a standardized facial recognition template provides another benefit – algorithms can use template extraction and matching from different vendors and still get an accurate identity check. This can prevent companies from getting locked in by a single vendor into a solution that doesn’t fit their use case.
The three essential edges of identity management are reliability, speed, and security. Thanks to ever-improving technologies, we are currently creating quantum leaps in all three – jumping closer and closer to secure digital and physical environments.
What are your considerations for the right cloud-based biometrics tool? Tell us on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We’d love to hear from you!