Facial Recognition Vs Facial Authentication
Vannessa Rataj
Face Authentication works by utilizing biometrics detection technologies. Biometrics are biological measurements or physical features that are used to identify individuals. Some common forms of biometrics include fingerprint mapping, retina scans, and facial biometrics.
During the face analysis phase, an image of the face is captured and analyzed by the device utilized. Again this image is converted into a faceprint by an algorithmic software. The identified facial biometrics in the faceprint allow the matching phase to be easier.
During the matching phase, the acquired faceprint is compared to the unique faceprint acquired initially for a correct match. AI and machine learning technologies may also be utilized for facial recognition, where they are trained with image recognition algorithms and training sets so that they can match a faceprint with different poses and environmental conditions.
The primary target of facial authentication is to utilize facial recognition technologies to identify correct individuals. As such, face authentication technologies have multiple applications for commercial ventures, entertainment and security services.
Globally, security services utilize face authentication to allow access to persons to critical areas and services within private companies or governmental institutions. Mobile face recognition has been utilized by law enforcers to use portable smart devices such as mobile phones to take images of a person to compare it to face recognition databases for accurate identification.
In airports, border control utilizes face authentication technology in conjunction with biometric passports to allow people to skip extended waiting lines in favor of an automated fate, where the face of the traveler is matched with their passport image.
Clients of banks can opt into facial authentication so that they can authorize transactions by looking at their smartphones or laptops with cameras instead of using OTPs on their devices. This adds another level of security in case the device in question is stolen as a second authentication factor.
Users can record their faceprint on their smartphones as a security feature to make sure only they can use the device in question. Some hotels also utilize face authentication to allow their guests access to their rooms or other amenities.
Social media platforms such as Snapchat and Instagram can utilize facial recognition for entertainment. These platforms can recognize faces for instant tagging or for using filters for people to enjoy.
Facial recognition is the technology capable of identifying or verifying an individual through an image, video, or any audiovisual element of their face. Facial Biometric Authentication utilizes this technology to allow an individual to access an application, service, or system.
Biometric identification uses measurements from bodies. In facial biometric authentication, the technology utilizes the face and head of an individual to verify the identity of the individual via their facial biometric pattern by collecting the unique biometric data associated with their face and expression.
Face Authentication is utilized when a 1:1 match is needed to access an application, device, service or system, while face recognition is utilized when a 1:n match type is needed to identify an individual. Face authentication matches the acquired facial biometric data to a specific user. Facial recognition matches the acquired faceprint to a database of faceprints, usually to identify the individual.
With the use of AI and machine learning, face authentication can be utilized with reliability and high safety standards where integration of algorithms and computing allows the process to be completed in real-time, further reducing any potential risks.
Facial authentication offers several benefits to their users, beginning with a very fast process. Facial recognition usually takes only a brief time and this speeds up the access process. This fast process allows for a unique and smooth user experience, where simply looking at a device is enough. Security is another benefit, as like fingerprints and voice, each face is unique. Facial recognition systems, especially with the assistance of AI, can compare facial biometrics very quickly. Compliance is another benefit, as face authentication through video identification is commonly the only method recognized as an industry standard for remote identity verification during high risk financial operations.
In this post, we will compare and contrast facial recognition and facial authentication and provide some cautions to companies considering facial recognition systems in commercial use cases when facial authentication is more appropriate, reliable and secure.
Apple deservedly gets much of the credit for making this technology available to common people in day-to-day life and for elevating the face as the most reliable, natural and secure biometric compared to its predecessors like fingerprint, IRIS sensors and voice recognition.
But, the use of facial recognition in real-time goes well beyond unlocking your phone. In fact, facial recognition technology is being leveraged within the United States and around the world, mostly for security purposes, and is increasingly being adopted for a wide range of applications, including:
Facial recognition is a way of recognizing a human face through technology. Modern facial recognition software use biometrics to map facial features from a photograph or video. It then compares the information with a database of known faces to find a match.
The facial signature (which is really just a mathematical algorithm) is compared to a database of known faces. This is what is known as 1:n matching (where n equals the number of face signatures in a database).
Facial recognition is ideal for matching a photo against large datasets of other facial images or videos, or comparing a high-quality selfie against another high-quality selfie. But facial recognition is not ideal for other use cases where a more definitive match is required. These scenarios include:
Facial authentication is the process of determining whether someone or something is, in fact, who or what it declares itself to be. After an online user has been vetted through some identity proofing or verification process, they are usually given a unique credential (i.e., username and password combination) that lets the user access their online account or perform certain actions. The verification of those credentials is what we call authentication.
Unlike facial recognition which performs a 1:n match against a database of known faces, facial authentication is 1:1. The user is authenticating using their face as their credential to secure access to their online account. To authenticate, the user simply takes a selfie, from which a biometric template is created and compared, one-to-one, with the stored biometric template. A proper match, based on an accuracy score, completes the secure authentication process in the background.
When the user needs to log into their account or perform a high-risk transaction (e.g., a wire transfer or password reset), the user is asked to retake a selfie. A new biometric template is created and compared with the one created during the initial enrollment and a match/no match decision is made in seconds.
Better facial authentication solutions will learn from each authentication event in a process known as adaptive learning. This means the new biometric template is compared not only to the original face map but all subsequent face maps to improve authentication accuracy and reliability.
Unfortunately, traditional authentication methods, such as SMS-based two-factor authentication and knowledge-based authentication, are no longer considered best practices because of reliability and security concerns such as phishing attacks and man-in-the-browser exploits.
By requiring a valid government-issued ID and matching it to a selfie (with certified liveness detection), enterprises can have a much higher level of assurance that the person is who they claim to be.
Of course, no technology is entirely without risk. Facial recognition is highly data-intense, which can make processing and storage an obstacle. Despite enormous advances, recognizing faces from multiple camera angles or with obstructions (such as hats) is still not perfect. Plus, there have been controversies related to privacy issues, particularly in retail and government settings. This is why facial recognition should not be used for identity proofing, but instead leveraged with multifactor methods (i.e., facial authentication) to strengthen user access.
The GaussianFace algorithm developed in 2014 by researchers at The Chinese University of Hong Kong achieved facial identification scores of 98.52% compared with the 97.53% achieved by humans. An excellent rating, despite weaknesses regarding memory capacity required and calculation times.
In June 2015, Google went one better with FaceNet. On the widely used Labeled Faces in the Wild (LFW) dataset, FaceNet achieved a new record accuracy of 99.63% (0.9963 0.0009).
Using an artificial neural network and a new algorithm, the company from Mountain View has managed to link a face to its owner with almost perfect results.
In May 2018, Ars Technica reported that Amazon is already actively promoting its cloud-based face recognition service named, Rekognition, to law enforcement agencies. The solution could recognize as many as 100 people in a single image and perform face matches against databases containing tens of millions of faces.
At the end of May 2018, the US Homeland Security Science and Technology Directorate published the results of sponsored tests at the Maryland Test Facility (MdTF). These real-life tests measured the performance of 12 face recognition systems in a corridor measuring 2 m by 2.5 m.
Thales' solution utilizing Facial recognition software (LFIS) achieved excellent results with a face acquisition rate of 99.44% in less than 5 seconds (against an average of 68%), a Vendor True Identification Rate of 98% in less than 5 seconds compared with an average 66%. It also achieved an error rate of 1% compared with an average of 32%.
c80f0f1006