Retail Checkout

Automated retail checkout systems aim to improve efficiency and reduce labor costs by automatically identifying and counting products during the checkout process. Current research heavily utilizes deep learning, particularly object detection models like YOLO variants and Vision Transformers, often enhanced with techniques like region-based approaches and digital twin training to address challenges such as occlusion, product similarity, and domain adaptation between synthetic training and real-world data. These advancements are crucial for achieving human-level accuracy in automated checkout, impacting both the retail industry through increased efficiency and the computer vision field by driving innovation in object detection and tracking algorithms.