Case study card:
Store planogram compliance automated with computer vision
Retail store audits is a common process used to assess the price and product compliance in the FMCG & CPG industry. Store audits usually answer questions like:
- Are products being displayed in the correct location in the store?
- Do the products displayed on shelves have mandatory SKUs, and are there any foreign
- Is the pricing of products and promotions in line with the company guidelines?
Generally, field surveyors conduct regular audits by observing the store and interacting with the staff. This is time-consuming and prone to error and corruption. Therefore, it is significant to find a way to make the whole process more objective.
By using computer vision, we have created models to detect products, location in-store, their signing and compare them with requirements stored in a database. This makes the whole process more objective as it no longer relies on human observation alone. This capability is served in a simple web UI which can be used by the field managers opening their stores as well as the managers to view the results.
— Azure Blob Storage with Blobfuse, Azure Virtual Machine, Docker, PyTorch, KP2D architecture, YOLO architecture