Researchers have developed a flexible machine learning system that can independently recognize products in store checkout or self-service systems. They published their results in the journal IEEE Access. Called PseudoAugment, it simplifies the self-checkout process and can be configured to recognize new products before they hit store shelves.
A group of researchers from Skoltech led by Andrey Somov created this neural network. It analyzes the images and looks for fake objects on them, structures similar to individual products. These objects are then retrieved and modified, thus creating many different pseudo-objects. This approach trains the neural network to recognize objects of different angles, sizes, and shapes even based on a small number of images, making it superior to its competitors.
The researchers tested the system on a prototype self-service checkout with a computer system to train scales and neural networks. Using just a few apple crate photos, they successfully trained the system to recognize new apple varieties with 92% accuracy. This algorithm can be applied not only in supermarkets, but also in other areas, such as the separation of seeds or municipal solid waste on conveyors, which can increase the efficiency of industrial plants.
Source: Ferra
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
