Retail theft costs A$9 billion annually. We developed a theft detection solution for a leading Australian supermarket chain
What we were solving for
Costs of retail theft: Analysing live footage across multiple cameras is challenging for staff, resulting in most instances of theft being undetectedAnalysing live footage across multiple cameras is challenging for staff, resulting in most instances of theft being undetected
Security resourcing: Utilising AI/ML for theft detection, reduces security staff resourcing requirements
OptiSol solution
Movements of all shoppers' within the store are tracked using object detection models like YOLO and tracking algorithm like DeepSORT
Utilising multithreading to process information from multiple cameras, the frames are retried in real-time and processed
A score based on suspicious activities is calculated for every shopper
Security staff are notified of suspicious activity via a web based portal