Bottom-of-basket (BoB) items in a shopping cart is a major source of loss of revenue in retail and this could run into billions of dollars.
The need for a solution to minimise BoB loss through Deep Learning by identifying if the cart is empty or not. TensorFlow was used as AI library, Visual Geometry Group (VGG) 16 Pre-trained on ImageNet database and Convolutional Neural Network was used to build at least 96% accuracy.
Tighter shrink control translated into better annual profits up to $60,000 per departmental store. Front end productivity also increased as transactions moved faster through the lane.