Facial recognition works through biometric technology, which recognizes a person by a unique and non-transferable trait. This automated identification mechanism, used for security and control, has become the most popular at the moment since it is in charge of measuring and analyzing all the biologic data from persons. Through facial recognition, it is easy to identify a person by analyzing the biometric characteristics of the face.

The facial recognition software works through a few stages: first, it detects or captures the face of the person that is going to be identified in the selected device, followed by the extraction of the facial characteristics, where the program obtains the biometric information of the facial traits. Next, it compares the obtained information with the one in the database. Finally, it obtains a percentage of the similarity of the person to be identified, and then it takes into account, to identify or not the face with the greater percentage of similarity, from an elevated threshold of consciousness. 

Amazon has a product called Amazon Rekognition, a software that facilitates the incorporation of the analysis of images and videos in applications.  This tool analyzes the images, through a line around the face and a square, that determines the only part of the image that Rekognition takes into account for the analysis. This technology compares data from the source image with every image found looking to give each face of the image a similarity score. Amazon Rekognition does not have information about the identity of a person, only the probability of the face matching to another person. 

The similarity score is a statistical measure of the probability that two faces of an image belong to the same person when the software analyzes it. The higher the similarity score is, it means that it is more likely that the two faces belong to the same person. Rekognition uses a system of probability, for this reason, there is not absolute precision, since the result is a prediction. 

Other methods for biometric verification 

The fingerprint identification is another way of biometric verification and this method is the oldest used to recognize a person, due to the different coating patterns in the fingers of each person. The technology does a thorough study of every fingerprint so that it matches all the traits and uniqueness of it. Also, there is a biometric through measuring the geometry of the hands; one step further from fingerprints since this method takes into account the different dimensions of the palm.

Another biometric is done by voice recognition, a system that identifies the person through voice patterns. This method is highly accepted by the clients to access a service and it is easy to implement, however, counts with a low resistance of deception since the basic parameters of the voice can be easily altered.

How can facial recognition contribute to security?

Through facial recognition, e-commerce can provide a higher level of comfort and security to the clients, since it would help them to keep their identity, instead of remembering passwords, codes, and security questions, or on occasions show documents of who they are. 

The biometrics do not require the clients to establish passwords that they have to keep remembering and hiding from everyone else, in contrast with the conventional processes of authentication. These are some benefits based on the increase of security when biometric authentication is used:

• Guarantees the clients secured access to the online store.
• The client will not have to worry about forgetting the password.
• Counterfeiting will be hard to achieve since the security measures will be reinforced.
• The identification of the users is immediate. 
• The accounts of the users will be non-transferable.

What are the main concerns?

One of the main concerns for e-commerce businesses when using this type of technology is how willing is the customer to move away from anonymity and expose his or her privacy? The reality is that using facial recognition is not recent, but implementing payments through this technology, although will avoid using credit or debit cards and other payment methods, will also be useful to get to know the demographics of the client. Characteristics like gender, age, facial traits, skin color among others, could be exposed and can be transformed into valuable data for the business, data that can be used to analyze the consumption patterns of the clients, to finally create content and offer custom made shopping experiences that are fuller and efficient. 

Conclusion

The experts believe that due to the comfort and the higher efficiency of this technology, the inclusion of facial recognition will increase for a secured shopping experience, for example, overcoming the geographical frontiers in the near future. The truth is that this e-commerce trend should be considered as long as it is easier to use without leaving to the side the customer experience, which facial recognition delivers to the client by speed and customization.

E-commerce businesses should analyze the behavior of their users and how well they are adopting these new methods and technologies. The adjustments of the online store and the different payment methods depend on the customer’s behavior. If the trust of the user is solid why not provide an easy and quick shopping experience to keep building customer loyalty? Also, the user will always appreciate the customization, in the end, they are the fundamental pattern in any business. 

‍