Retail product shrinkage is a $100 billion per year challenge, with over 65% of shrinkage—and growing—due to theft. With NVIDIA’s retail loss prevention AI workflow, you can quickly build and deploy applications designed to prevent theft. Leveraging the NVIDIA Metropolis microservices and models pretrained on hundreds of the most frequently stolen goods, this AI workflow provides few-shot, active learning to quickly scale to recognize hundreds of thousands of products and provide intelligent alerts with actionable information.
Explore the Retail Loss Prevention AI Workflow
Built on cloud-native NVIDIA Metropolis microservices—a low- or no-code way to build AI applications—this reference workflow for retail loss prevention and frictionless shopping delivers pretrained AI models along with the applications needed to jump-start development and rapidly index hundreds of thousands of store products for cross-camera and barcode-scan identification.
The AI models in this workflow are pretrained to recognize the hundreds of products most frequently lost to theft—including meat, alcohol, and laundry detergent—and to identify them in varying sizes and shapes. With synthetic data generation from NVIDIA Omniverse™, retailers and independent software vendors can customize and further train the models to cover hundreds of thousands of store products. The workflow is based on a state-of-the-art, few-shot learning technique developed by NVIDIA Research that, combined with active learning, identifies and captures any new products scanned by customers and sales associates during checkout to ultimately improve model accuracy.