20. February 2023
In the modern digital world, it has become easier than ever—and in some respects, more necessary than ever—to collect information on consumer shopping habits, preferences, and behaviors. At the same time, data protection standards such as the GDPR oblige retailers to maintain the privacy of their customers. And that’s where anonymization solution for retailers comes in.
“As retailers increasingly turn to computer vision-powered technologies such as self-checkout, inventory management, and video surveillance, we can expect to see an exponential growth in the amount of data being generated.”
Andreea Mandeal, Head of Marketing, brighter AI
According to the 29th Annual Retail Technology Study by RIS, only 3% of retailers have implemented computer vision technology. Yet 40% are planning to deploy it within the next two years – for a good reason.
Buying into computer vision technology
Computer vision helps retailers to improve their performance by creating frictionless self-checkout experiences, addressing pain points, optimizing store layouts, and transforming the customer experience. Virtual mirrors allow physical stores to offer online levels of personalization and convenience. And all kinds of routine tasks can be automated, freeing employees to spend more time on customer-focused activities.
5 ways computer vision enhances the retail experience
- Video surveillance
When we talk of computer vision, video surveillance is the use case that most commonly springs to mind. The advanced capabilities of machine learning and computer vision allow retailers to automatically identify suspicious patterns of behavior, catch shoplifters in action, and alert security staff.
- Self-checkout
As things stand, self-checkout usually involves shoppers scanning the barcodes of the items they place in their baskets. Computer vision-powered cameras take this to the next level by automatically recognizing products, making the entire experience much faster, far easier, and more secure. Amazon’s Just Walk Out system is a great example: combining cameras, sensors, and deep learning it allows customers to simply pick up the items they need and walk out of the store. No scanning. No waiting in lines. And no app required.
- Inventory management
Computer vision is also being deployed to optimize retail inventory management. Shelfie is just one example of many, with cameras mounted above shelves alerting staff to damaged packaging, out-of-stock items, or products placed in the wrong section.
- Store layout optimization
Retailers can also use computer vision cameras to track customer movements and identify their purchase patterns. These insights are then used to place products, distribute staff, and even lay out the store in the optimal way.
- Virtual mirrors and recommendation engines
Virtual mirrors may well represent the next generation of personalization and customer experience in the retail environment. A virtual mirror is essentially a conventional mirror equipped with computer vision cameras, AR, and a built-in display – such as FindMine’s ‘Complete the Look’ fitting room technology. Say a customer tries on a pair of trousers, the virtual mirror will recognize the item and recommend matching pieces to complete the look.
With lots of data, computer vision will transform the retail experience
So computer vision is set to transform the physical retail experience. Yet vast amounts of video data are required to train the machine learning algorithms at the heart of these technologies. Remember those video surveillance cameras we described? According to the CEO of tech startup Vaak, it takes 100,000 hours of video data to train their systems. Which brings us to the importance of anonymization solution for retailers. Because it’s not just a matter of reassuring the many customers who feel uncomfortable in having their movements tracked by cameras. It’s also a matter of complying with the law.
Robust international data protection laws place retailers under a legal obligation to protect the personal information of their shoppers. This leaves them facing a challenge: how to reap the benefits of computer vision technology powered by data-driven AI algorithms, while at the same time safeguarding the data of their customers. The answer to getting the balance right lies with anonymization.
The role of anonymization and DNAT
Anonymization is a technology that allows retailers to collect useful information about the behavior of consumers, while still protecting their privacy. This involves making sure that all faces and any other recognizable data are rendered completely anonymous, so there would be no way of identifying any subject from the images. The trouble is, traditional techniques such as pixelation and black bars are unable to preserve the accuracy and integrity of the original data, rendering it is ineffective for machine learning.
Deep Natural Anonymization (DNAT) resolves this trade-off between innovation and privacy. DNAT uses generative AI to create synthetic faces and replica license plates and prevent the original subjects from being recognized. At the same time, it preserves the quality and integrity of the original data. In doing so, it represents the only anonymization technique capable of powering analytics and machine learning.
So why should retailers care about anonymization?
In short, it’s good for business. Anonymization allows retailers to make the most of next-generation computer vision technologies to run their stores more efficiently, boost sales, and create stronger connections with customers – all while complying with ever more stringent protection laws. Failure to do so not only threatens to cause major reputational damage, but can also lead to equally hefty fines. In fact, 43% of UK retailers have been fined for their failure to protect customer and employee privacy in video footage. (Source: The Retail Bulletin)
As the world becomes increasingly digital, video-powered retail technology is quickly gaining traction as retailers start to understand the benefits on offer. At the same time, shoppers care about their privacy, while regulators are ready to clamp down hard on any retailer that violates local regulations. By using anonymization solution for retailers, they can collect the essential data they need to stay ahead of the curve, while keeping that data safe from misuse or theft.
