Latest News

5 inventive uses of Computer Vision in retail

Manu Krishna, Associate Director, Marketing, at Trax discusses how vision powered technology is supporting retailers


Early research on Computer Vision started over 50 years ago, and its application across industries have grown in sophistication along with our understanding of the discipline. Most digital cameras today recognise faces in a picture, OCR software in scanners converts scanned documents to text, and vision-based biometrics also famously helped identify an Afghan girl by her iris patterns.

One of the most exciting areas we are seeing this technology being used is in retail. While early applications of Computer Vision in retail come from e-commerce, increasingly it is being used in physical retail stores to enhance operational efficiencies and create a frictionless experience for shoppers.

In recent years, brands and retailers have become more inventive in their use of Computer Vision and have excelled in innovative services and solutions using vision-powered technology.


Blurring the line between in-store and online

Sometimes a consumer see something they want to buy, but have no information about it. That’s why a tool called Lens has been created. It’s been launched by the photo sharing website Pinterest as a beta product, and is set to improve the in-store experience for shoppers everywhere.

It works by recognising an object and providing contextual information about it. It can tell you anything from who designed a piece of furniture and in what year, to what other clothing in the store will go with the pair of shoes in front of the shopper. Consumers just take a photo of the item, and the app does the rest.


Identifying regulars, rewarding loyalty

Gourmet candy retailer Lolli & Pops uses facial recognition to identify loyalty members as they walk into the store. Computer Vision then enables a personalised shopping experience: by scouring shoppers’ purchasing history and preferences, the system can make personalised product recommendations specific to each shopper.

By treating them as individuals – and more importantly, as VIPs – the system instills brand loyalty, and converts occasional shoppers into regular customers. Both of which are good for business.


Transform the health of every shelf, in every store

The beauty and simplicity of Computer Vision is its ability to turn actual images into actionable insights in order to help brands and retailers focus on fundamentals in the store. By digitising the shelf, companies now get real-time situational awareness about what’s happening on the shelf. The directives range from the obvious – such as: “go to the back room and get a box of product to fill an empty space” – to the more sublime, such as instructions to reduce the number of facings (how many products of the same type that are sitting side-by-side) of a competitor and increase your own facings by that same amount.

Non-mobile users can also get role-based insights on a huge array of retail metrics that tell them exactly what’s happening on-shelf and what to do to ensure the best shopping experience, and ultimately drive better sales.


Analysing footfall, pass-by traffic, interactions and more

Aurora by RetailNext is the first sensor designed specifically to meet retail’s complex needs. It counts in-store footfall traffic like so many other sensors of its ilk, but also adds texture to the data – it includes the capture rate of pass-by traffic, and breaks down shoppers’ paths around the store. With this data, retailers can see which promotions capture engagement, and which turn customers off.

But it doesn’t just monitor the shoppers. It also adds customer and associate interaction, providing real-time visibility into in-store service engagement. Plus it can be used to drive personalised marketing and messaging campaigns.


The frictionless store experience

Computer Vision can also help when it comes to one of the worst parts of the shopping experience: queuing at the tills.

The Amazon Go concept store in Seattle tracks shoppers using Computer Vision, with sensors on the shelves detecting when they pick up an item. It then registers all the items in the shopper’s shopping basket with the Go mobile app, and removes the checkout process altogether – the shopper simply leaves the shop, with the Go app taking the money automatically from the shopper’s nominated credit card. The receipt is sent straight to the app.

The ever-connected shopper experiencing frictionless retail is truly where the retail industry is headed, something that is only made possible by a combination of Computer Vision and deep learning.

error: Content is protected !!
Don`t copy text!