Abstract
Reporters obtained the following quote from the background information supplied by the inventors:“Automated checkout systems allow a customer at a brick-and-mo rtar store to select and purchase itemsall without aid from a human attendant. These automated checkout systems may automatically identifyitems that a user co llects as the user collects them. For example, the automated checkout system mayidentify items in images captured by the user’s client device or by cameras att ached to a shopping cartto identify items that the user has collected. In some cases, the automated checkout systems may usemachine-learning models to identif y the collected items. For example, the machine-learning models maybe trained t o identify an item based on a machine-readable label (e.g., a barcode or a QR co de) affixedto the item. U.S. patent application Ser. No. 17/703,076, filed Mar. 24, 2022, describes examplemachine-learning models for decoding machine-readab le labels based on images, the contents of whichare incorporated by reference.
NETL
NSTL