The Near future of retail billing — AI
For some it is an addiction, for some, it is a stress buster, for some it is the need that forces them to visit the store. But one thing that all offline shoppers agree upon is the long-long waiting at the bill counter to get their things billed.
Studies show that an average human spends 2 years of his life standing in billing queues [1].
Fortunately, Artificial intelligence (AI) has come to our rescue. AI being the dynamic technology, it is has come up with a solution to these long queues. Major companies are coming up with retail store ideas that completely run on AI.
Ok. But how does this technology work?
There are quite a handful of ways it can be done.
One among them is used at Amazon go. The is store located in downtown Seattle and completely runs on AI. They call it “just walk out technology” which uses computer vision, deep learning algorithms and sensor fusion. So the customer just scans the app, enters the store, takes the product and just walks out. As simple as it sounds.
Prior to entering this store, the customer should download the app and scan the QR code at the entrance. It works as an identity for tracking the movement of the customer when he/she is in the store. The sensor fusion might include the weight sensors placed below the product to know which product is chosen by the customer and when is it replaced by the customer.
Apart from such sensors there are also hundreds of cameras fixed to the ceiling of the store to track the movements of the person so that it can identify the person, the product chosen by person to add the products to the person’s virtual cart. When the customer walks out of the store with the products amazon charges his/her account and sends a receipt to the customer.
Where does Annotation fit in this case?
As we know the system, with the help of weight sensors, recognises when a product is picked. But how does it know which product is picked?
Annotation is the process that helps the system to know which product is chosen by the customer. Firstly, a label is given to the product for classifying the product and then the image is annotated and fed to the system for product recognition. It is not just the product but also the customer is annotated to know who is picking which product and who is replacing what. As these products and human do not have a definite shape, the polygon tool can be used for annotation as it gives a precise image for training the system.
What is OCLAVI?
OCLAVI being a platform that provides benefits like
· Annotation of images with the low-quality range.
· The Option of working with their own team or with a freelancer
to its clients, can make machine learning far better than before.
In this way, AI can bring in many innovations into the retail business giving it a more technological perception to not just save time but also for a better shopping experience.
Explore more OCLAVI
References :
