Download our Rapid Image Recognition White Paper.
Download our Rapid Image Recognition Fact Sheet.
Watch our Webinar “Introducing Exceedra’s ‘Next Generation’ Rapid Image Recognition”.
On this week’s episode of Exceedra Byte, Andres Jejen, Exceedra VP of Solutions Engineering, explores one of the most common applications of Machine Learning in Retail Execution: Image Recognition.
For those that are not familiar with the functionality, through image recognition it is possible to take one or more photos of a product location (such as a shelf, an endcap, a fridge, etc.) and to obtain identification of the products found there. The key to this technology is not only to identify what products were found, but to be able to provide insights on the condition of the products in that location in an automated way.
The automation of the data acquisition at the stores brings several benefits around revenue increases. For instance, you can reduce your in-store time significantly, up to 80%, compared to the time spent doing manual shelf counts.
On the other hand, with the time savings, you can increase your coverage by adding new stores per day or you can increase your visit effectiveness by dedicating more time to other productive activities. By honing into those factors, you can see sales increases and expense reductions in a relatively short time.
You might be asking the question – if this is so great, why is that it is not massively used?
Although this technology is used relatively widely, the truth is that several factors have prevented it from being massively and completely adopted. Things like cost, complexity and availability in some markets have impacted the ability of companies to embrace this technology as their new game changer. But solution providers like Exceedra have been working to address those factors, and now you can implement Image Recognition at an affordable, simple, and effective way and you can start reaping benefits for your business very quickly.
Sounds great, right? But you still don not think this is for you? You have so many packaging variations, reflecting materials, odd-shaped products that you do not think the software can help you out here.
The way this technology works follows a very simple principle: if through the set of images, the human eye can determine with accuracy which products are there, the machine will do it as well but at a higher level of speed and precision. In that sense, it does not matter if your product comes in bags, cans, bottles, or boxes. On the other hand, if your packaging changes seasonally or recurrently the system can be easily enhanced to identify the variations and, in some cases, even to continue recognizing the product as is, since the algorithm is based on brand elements, such as logos, color schemes, curves and any other discernible graphical asset.
But wait, there is more! Image Recognition can do much more than just simply identifying products on a location. Exceedra’s own Rapid Image Recognition is tailored to do measures of adjacency, relative positioning and even a complete display compliance, by analyzing presence of both products and promotional elements at once. Also, price tag presence and contents can be read and used as additional data points for the analysis.
If you are interested in testing this technology to see how beneficial it can be, reach out! We can have a pilot up and running for you in no time.
BLOG Localizing assortments in an era of lumpy supply chains RGM Series Consumers still want to shop local post pandemic. How to focus on ranging loca...