Face recognition is not a new phenomenon – integration of traditional face recognition systems with the power of Artificial Intelligence is a revolution. Just think how social networking apps like Instagram can map your face’s shape accurately. The user ID adapts to changes according to your look – glasses, makeup, facial hair, etc. The smart system design is compatible with all environments – outdoor, indoor, low light conditions and even darkness.
Working Principle
A facial recognition system identifies or verifies a person’s identity by analyzing their face. The raw input for the camera is the person’s face recorded in real-time. The AI-enabled face recognition system captures the person’s image from the recorded video, analyzes it and compares it with images stored in its database. Facial recognition combined with a biometric fingerprint is useful for access control and prohibition of entry for unintended persons.
Benefits
Primarily, a face recognition camera involves front-end operations. Popular analytics include resource distribution control, police verification, facial similarity, human trajectory, key reminders, cost efficiency, etc.
In the real estate industry, builders deploy these cameras at the entrance/exit of apartments. It helps to identify proprietors and enables real estate companies to address trade barriers. The finance industry businesses are installing face recognition cameras to identify customers for designing membership plans, customized Sales/Marketing strategies, and create efficient Sales opportunities.
Let us look at the example of the Tend Secure Lynx. This smart camera not only performs well but also comes with several features – it stores video clips event-wise over a period of 7 days and that too free of cost. Users have to just create the database of faces familiar to the user. The Lynx will take some time to learn about the faces but once it is familiar, the camera is useful for indoor home security.
Loopholes
Every technology has its own advantages and disadvantages. Cybercrime is a real problem. Technology enthusiasts have identified the loopholes of the AI Face Recognition camera. They have come up with a system to confuse it.
Grigory Bakunov, a Russian technologist has come up with a solution that escapes the camera and cheats face detection systems. The algorithm creates artificial makeup which tricks the intelligent software. The product has not been launched in the market to avoid cybercrime.
In Berlin, Germany Adam Harvey, a professional artist has invented similar equipment called CV Dazzle. Currently, he is focusing on clothes, as he is working with apparel having patterns to avoid detection. The disguise comprises of fabric patterns – eyes, mouth, etc, to confuse the face recognition camera.
Towards the end of 2017, a company from Vietnam hacked the Face ID function of iPhone X by using a mask. It is a complex hack and cannot be used for amass exploitation.
Consequently, German researchers designed a hack for Windows 10 Hello. The users could bypass facial authentication through printing an infrared facial image.
A user can apply filters to modify the image resolution before uploading it on the website. These changes are evident for the human eye but escape intelligent algorithms of the face recognition camera.
Read more: Why Facial Recognition Providers Must Take Consumer Privacy Seriously
Comments are closed, but trackbacks and pingbacks are open.