Abstract: Embodiments herein describe providing feedback to a shopper at a POS system using a computer vision system. Many items at a store may lack barcodes or other identifying marks such as produce. The shopper may have to perform an action to identify the item to the POS system. The computer vision system can double check the identify provided by the shopper to reduce mistakes and deter nefarious actors. If the computer vision system cannot independently confirm that the item being purchased matched the identity provided by the shopper, the POS system can display a graphical user interface (GUI) that includes an image of the item captured by the computer vision system along with identification data of the item identified by the shopper. This gives the shopper a chance to correct any mistakes.

Abstract: Techniques for training an item recognition machine learning (ML) model are disclosed. An image of a first item for purchase is received. The image is captured by a point of sale (POS) system. A purchaser selection of a second item for purchase is also received. The purchaser makes the selection at the POS system. It is determined that the first item for purchase matches the second item for purchase, and in response training data is generated for an image recognition ML model, based on the image and the determination that the first item for purchase matches the second item for purchase. The ML model is trained using the training data, and the trained ML model is configured to recognize items for purchase in a plurality of images captured by a plurality of POS systems.

Abstract: Embodiments herein describe providing feedback to a shopper at a POS system using a computer vision system. Many items at a store may lack barcodes or other identifying marks such as produce. The shopper may have to perform an action to identify the item to the POS system. The computer vision system can double check the identify provided by the shopper to reduce mistakes and deter nefarious actors. If the computer vision system cannot independently confirm that the item being purchased matched the identity provided by the shopper, the POS system can display a graphical user interface (GUI) that includes an image of the item captured by the computer vision system along with identification data of the item identified by the shopper. This gives the shopper a chance to correct any mistakes.

Abstract: Techniques for automatic recognition of items for point of sale (POS) systems are disclosed. A request to identify a first item for purchase is received. The request includes a first image captured at a POS system. The first image is analyzed using a machine learning model configured for image recognition, and in response a first product code is determined for the first item. This includes identifying the first product code as a primary product code from among a plurality of product codes, identifying a second item relating to the first item, where the second item is visually similar to the first item, and determining a second product code for the second item. The first product code, and the second product code, are transmitted to the POS system. The POS system is configured to present the first item and the second item as options for purchase using a user interface.