Face recognition: criteria for choosing a popular biometric technology

розпізнавання облич
calendar05 August 2022

Did you know that people can only recognize faces with 85% accuracy? Of course, when a person begins to speak or an expression appears on his face, it is much easier to recognize him.

Modern video analytics systems have recognition accuracy of up to 95-98%, which contributes to the use of technology in many areas. Conducting investigations and preventing crimes, creating a portrait of the target audience, access control and accounting of working hours are just a small list of tasks where face recognition technology is used.

In addition to accuracy, what other parameters should you pay attention to choose a solution?
Recognition by additional attributes
Glasses, facial tilt, mustache, beard, headdress or medical mask should not interfere with face recognition.

Determination of descriptive characteristics of a person
Gender, age, race, emotions of a person, color of clothes, length of hair – this is an incomplete list of what the system can detect.

Formation of lists
Create an unlimited number of checklists of people: visitors, staff, VIPs or undesirables.

Identification speed
Modern video analytics systems allow you to recognize up to 200 people in a frame at the same time and identify them in a matter of seconds.

Event notification format
How convenient will it be to receive messages? It can be an e-mail, a notification in Telegram or an SMS message.

Flexibility and effectiveness
Statistics and reports should ensure the fulfillment of analytical goals, calculation and evaluation of key indicators. The solution should allow you to build your own business logic, have wide opportunities for integration with other systems and products.

Of course, intelligent face recognition systems still have a long list of technical criteria to choose from. Contact INLIMITED, we work with Ukrainian developers of video analytics systems and will help you in selection and testing of solutions.