With 15,000 visitors per day, and a watch list of about 3,000 people, the casino faced an impossible task to recognise persons of interest in real time. Surveillance staff manually detected only 5–7 known persons per week and was looking for an automated face recognition system to support two separate requirements:
