With the development of computer vision technology that analyzes images containing faces, there have been many questions about the security of doing so. However, if people understand the differences between the two primary applications: facial detection and facial recognition, they can eliminate some of the concerns.
Facial Recognition System
A facial recognition system is designed to answer the question: can the face in this image be matched to a face in another picture? The key idea to a facial recognition system is to compare the face of an image to images in a database of stored records and establish whose fact it is.
Some people may accept the idea that companies can use this biometric technology to enhance their customer experience. For example, a coffee shop may quickly prepare a cup of coffee and a piece of croissant when the shop can identify that a customer with such regular order walks into the store. Other people, on the other hand, may feel uncomfortable about companies having access to biometric data that can be traced back to them.
Facial Detection System
The technology that can satisfy these two groups of people at the extremes is facial detection. Instead of identifying a specific person from the face as a facial recognition system does, a facial detection system is designed to answer the question: is a face present in this image? The facial detection system can only tell whether there is a face in the image and identify general information about a person without identifying who that person is.
Since a facial detection system does not involve comparing incoming images to images in the database, there is no need to store any personal biometric data. Without any data storage, people do not have to worry about data breaches. While relieving concerns regarding data security, the facial detection system also enables retail store owners to provide better customer experience to their customers. A facial detection system can identify general information about the face, such as age or gender, without specifically identifying who the person is. If retailers discover that the majority of customers making purchases at their stores is from a specific demographic group, they can optimize the product choice and marketing campaigns to better target the demographic group.
This does not imply that facial recognition is an unacceptable technology. Facial recognition has its applications in areas where security is of utmost priority, such as in government settings. However, in settings where such specific recognition is not necessary, facial detection may be a good choice of technology without any concerns regarding data storage and security.
Cyclops uses facial detection to generate insights for retailers. Without having to trace the facial image back to a specific person, Cyclops can analyze the age, gender and facial expression of customers who visit the store. This information helps brick-and-mortar store owners to understand what kind of people are visiting their stores and how they feel during their in-store shopping journey.