TECHNOLOGY Facial Recognition Technology

How Facial Recognition Works

Former face recognition techniques were Voila-Jones Haar cascade or a histogram of oriented gradients. However, modern facial recognition technology is based on a specific neural network called convolutional neural network. To match the face templates, convolutional neural networks process each image through several steps:

Face Recognition Steps

1. Face detection: The first step is detecting a face within a larger image or scene. This process involves distinguishing facial features from the surrounding environment and pinpointing their location within the frame.
2. Face Analysis: Once a face is detected, the technology analyzes the facial features. This analysis is typically based on the geometry of the face, measuring various key points on the face, known as landmarks or nodal points, which can include the distance between the eyes, the shape of the jawline, and the contours of the cheekbones, lips, and nose.
3. Feature Extraction: The analysis results in the extraction of facial features, which are used to create a faceprint or face template—a digital map of the face’s geometry.
4. Comparison: This face template is then compared against a database of known faces. This is done using sophisticated matching algorithms that can handle variations in lighting, facial expressions, and angles.

A convolutional neural network converts every face pattern into a numerical code with every template expressed as a numerical vector. The closer two vectors are to each other, the more likely there is a face match between them.

Facial recognition technology operates by capturing a digital image or video frame that includes a face. This image is meticulously analyzed to identify specific landmarks or features on the face, known as nodal points. These nodal points, which are crucial elements of the individual’s facial geometry, such as the distance between the eyes or the width of the nose, are used to create a face template—a digital representation of the face’s unique characteristics.

Modern facial recognition systems employ sophisticated algorithms that can assess and log more than 80 such nodal points from the face template. The information from this face template is converted into a mathematical formula, resulting in what is termed a facial signature. This signature is a distinctive numerical code that encapsulates the facial features stored in the database. When facial recognition technology encounters a new image, it compares the new face template to the facial signatures within an existing database of known faces. Utilizing a complex array of artificial intelligence algorithms, the system evaluates the face template with impressive speed and accuracy to ascertain if there’s a match with any stored facial signature.

The key to the evolving success of facial recognition technology lies in advancements in machine learning, which refine the process of creating and comparing face templates. As the system learns from each face it processes, it incrementally increases its precision, reinforcing the technology’s pivotal role in areas such as security, personal authentication, and various applications across our digital ecosystem.

Core Technology and Accuracy of Facial Recognition

Facial recognition technology’s core is anchored in its sophisticated ability to discern and identify individual faces from digital images and videos. This sophisticated process encompasses several stages: detecting a face, analyzing key features, extracting these features to create a faceprint, and finally, comparing it against a facial database for recognition. Underlying this sequence are pivotal infrastructures such as advanced artificial intelligence (AI), leveraging machine learning and deep neural networks, integral computer vision techniques for high-level image processing, and biometric analysis for capturing unique facial characteristics. When integrated into B2B systems, this technology not only enhances security but also streamlines operations, offering seamless user experiences across various digital platforms.

Core Technologies

• Artificial Intelligence (AI): AI, particularly machine learning and deep learning algorithms, is at the heart of modern facial recognition systems. These algorithms enable the system to learn from vast amounts of data and improve over time.
• Computer Vision: This field of AI is concerned with how computers can gain high-level understanding from digital images or videos. It involves the automatic extraction, analysis, and understanding of useful information from a single image or a sequence of images.
• Biometric Analysis: Biometrics refers to the statistical analysis of biological data, which in the case of facial recognition is the unique features of the face.
• Neural Networks: Particularly, convolutional neural networks (CNNs) are used for image recognition tasks and are crucial for the feature extraction phase in facial recognition systems.

Together, these technologies form the bedrock of facial recognition systems, allowing them to function with increasing levels of accuracy and reliability.

Facial Recognition Accuracy Is Improving

Facial recognition algorithms are not as accurate as fingerprint or iris algorithms. However, given the development of convolutional neural network, the accuracy of facial recognition algorithms has been on the uptick.

The accuracy of each algorithm is measured by two error classes:

False Positivity

False Negativity

False positivity occurs when the software wrongly considers photos of two different individuals to be the same person. On the other hand, false negativity means the software failed to recognize the same person. There is always a tradeoff between false positivity and false negativity, therefore, it can sometimes be difficult to set up the thresholds in biometrics.

Face recognition accuracy is rapidly increasing. The world’s leading face recognition algorithms have reduced error rates by orders of magnitude (from 10% to below 1%) in just 3 years. Between  2014 and 2018, the accuracy of facial recognition technology increased 20-fold. These continued improvements over the next several years will likely bring a myriad of new use cases. See Innovatrics results here.

Using Biometric Templates on Personal Devices

Biometric authentication is now used on personal devices instead of PIN or built-in within application as a verification or client onboarding tool. However, the main reason mobile phone manufacturers use it is due to its convenience. There is no need for password protection and management. Also, both fingerprint and face recognition are much faster than traditional authentication methods. There is also more security, because no one can simply look over the shoulder to steal your password. Finally, as there is match-on-device technology and the data is never provided to third parties nor is it stored in the cloud, the user can be sure no one else is tampering with it.

Applications and Use Cases of Facial Recognition

Nowadays, facial recognition technology has a broad spectrum of applications worldwide – from individual onboarding and personal security to the identification of individuals in crowds and gatherings. The most common use cases of this technology include:

• Criminal identification of suspects
• Digital onboarding of clients
• Access control
• Security and surveillance
• Identity fraud prevention
• Event registrations
• Airport operations
• Financial services verification
• Smartphone unlocking
• Retail analytics