Opinions expressed by Digital Journal contributors are their own.
Facial Recognition (FR) technology has become increasingly prevalent across various industries in the modern era of rapid technological advancements. Its widespread usage in law enforcement, aviation, retail, and security sectors has significantly expedited the process of identifying individuals. However, while Facial Recognition technology has undoubtedly improved its efficiency and accuracy, it still faces certain limitations, especially in differentiating individuals with similar physical features.
These limitations highlight the importance of incorporating a human element in the review of FR results. The “human in the loop” concept involves using human practitioners and technology to increase accuracy and reduce potential errors. In the context of facial recognition technology, a human reviewer would verify the results of FR systems before taking any actions based on them.
In law enforcement, where the consequences of false identifications can be grave, the “human in the loop” approach is especially crucial. For example, a false identification could lead to wrongful arrests or convictions. By including human reviewers in the facial recognition process, law enforcement agencies can ensure the technology is used accurately and fairly. This approach allows for human judgment and expertise to supplement the capabilities of FR technology, helping to prevent potential errors and false identifications.
Ideal Innovations (I-3) fills this crucial gap with its specialized Facial Identification services within its Face Center of Excellence (FaCE). Established in 1998, I-3 is a technology-driven identity solutions and services company that excels in providing biometric and forensic services. In addition, their services include facial comparison consulting, expert witness testimony, and training for individuals to be the “human in the loop.” I-3’s services have been employed by various federal organizations, such as the Department of Defense and the Department of Homeland Security, as well as international entities, state and local agencies, private firms, and individuals.
It is essential for the “human in the loop” to be properly trained to minimize the probability of misidentification—proper training includes understanding the limitations of the facial recognition technology they use. In addition, knowing how the system works, what data it uses to make its identifications, and what factors can affect its accuracy is crucial in evaluating the results of the system and ensuring that they are accurate.
Additionally, training should include topics centered around working with low-quality images, images with poor lighting, or images with partial faces. Similarly, examiners should have training in factors that can affect identification accuracies, such as changes in facial features due to aging or injuries.
Finally, the human reviewer needs to be trained in ethical and legal considerations surrounding facial recognition technology. In addition, they must understand the potential consequences of a false identification, particularly in law enforcement.
A National Institute of Standards and Technology (NIST) study entitled “Face recognition accuracy of forensic examiners, superrecognizers, and face recognition algorithms” found that trained human examiners were more accurate than untrained persons. When the trained human examiner was fused with an automated Facial Recognition algorithm, the results were near 100% accuracy. This is because human examiners can consider factors such as lighting, angle, and other environmental variables that may affect the accuracy of Facial Recognition technology. It is important to note that proper training is crucial in ensuring the accuracy and reliability of the facial comparison process.
Facial Recognition technology has revolutionized the way we identify individuals. However, a trained human in the loop is necessary to ensure accuracy and reliability in the process. The I-3 Face Center of Excellence is committed to a trained human in the loop, which minimizes errors in the facial comparison process by providing training to those who need it and consulting services to those who don’t have staff to conduct comparisons. For more information on Ideal Innovations Inc., or to learn about the services they offer, visit their website at https://www.i3face.com/