The thesis, Barcodeless Food Products Recognition for Retail Self-checkout Service, was completed at the Vilnius University Institute of Data Science and Digital Technologies under the supervision of Prof. Dr. Povilas Treigys.
Research focused on a real retail problem
Fruit, vegetables, bakery products and other items without barcodes can be difficult to identify at self-checkout. Products may look similar, appear differently under changing light or be placed on the scale in different ways.
The research explored how AI systems can recognise these products accurately while running on computers with limited processing power. It also addressed practical situations such as products in bags, customers' hands appearing in the image and differences in product sales volumes.
Supporting Retellect's product development
The research is closely connected to the problem Retellect is solving for retailers. It strengthens the technical foundation behind the product recognition system and supports its development for real self-checkout environments.
For Retellect, the completed PhD marks another step in turning specialised research into technology that can be used in stores.