Airports, building security, health monitoring, and marketing and retail use facial recognition technology. Law enforcement agencies are also beginning to incorporate the technology in looking for investigative leads. Yet, despite its benefits and potentials, a host of problems with it are surfacing. For instance, concerns abound around data privacy, data protection, racial profiling, and human rights abuses. Recently, lawmakers have been responding to calls to issue a moratorium on the use of the technology. And they should be, because with the current level of technology, facial recognition technology is more of a problem than it’s worth.
Popular culture often depicts facial recognition technology as a sinister scheme to track and monitor the population. In the Ridley Scott classic “Blade Runner”, a retina scan machine is used to detect rogue Replicants. But perhaps with the developments in artificial intelligence, data analytics, and facial recognition technology, the movie “Minority Report” stands as a more apt portrayal of the world we are currently living in (with holographic ads keyed toward consumer data stored in eyes!). Certainly, facial recognition technology will change the way we live in the coming years. Thus, strengthens the need to understand and create mechanisms to regulate powerful surveillance and identification technology.
Applications of Facial Recognition Technology
Developed in the 1960s, the first face recognition algorithm used semi-automated technology. Facial ratios and calculated distances based on eyes, nose, mouth, and ears were used as reference points. For decades, military and law enforcement used facial recognition technology as a tool. However, with the emergence of new software, data capture systems, and a growing database of photos, it has found its way in a variety of applications.
Currently, usage of the technology includes unlocking phones and accessing e-wallets for digital payments. There is also facial recognition technology to enhance tenants’ security in buildings. The hospitality industry has also joined the foray. Some hotels have begun using facial recognition technology to greet guests and facilitate check-in. In health care, facial recognition technology has been used to monitor high-risk and ICU patients.
Gaps in Facial Recognition Technology
While it is true that the technology has been beneficial across industries, there are areas in the facial recognition technology that can be improved. As the application of the technology increases, gaps, and problems in the facial recognition technology need to be tackled immediately. Otherwise, the potential for misuse is high.
- The reliability of the technology is still under scrutiny. Currently, the technology has challenges identifying individuals of color. This flaw raises the concern around racial bias and neutrality.
- When asked for a match, systems return with several facial matches. Combining the inconsistent algorithm of facial recognition technology with human intervention, there is an increase in the risk of bias and subjective interpretation.
- Problems in facial recognition technology include the accuracy and quality of the database. There is no minimum photo quality. This means altered and edited photos can be added in the database, subsequently distorting the matches and results.
- The pace of technology assimilation give legislators and regulating bodies a hard time catching up. Without laws and regulations within the space keeping the actors in check, opportunities for misapplication of the technology abound.
False Positives and False Negatives: A Drill Down on Problems with Facial Recognition Technology
The last ten to fifteen years witnessed remarkable developments in facial recognition technology. Approaches and algorithms such as Principal Components Analysis (PCA), Linear Discriminant Analysis (LDA), and Elastic Bunch Graph Matching (EBGM) have been developed. The growing interest towards the commercial application of the facial recognition technology has paved for innovations around image capture, scanning, indexing, and data retrieval. Despite this progress, the reliability and accuracy of the technology are still uncertain.
False positives – the incorrect flagging of subjects – have been occurring. The wrongful arrest of eighteen-year-old Ousmane Bah from New York led him to launching a legal battle against Apple. He was flagged for robbery and theft. On the other hand, false negative happens when the system fails to find a match for a subject despite existing records in the database. Experts in the field are aware that the system has challenges recognizing women of color.
Efforts to address these problems are underway. However, these gaps can have a damaging and irreversible impact on peoples’ lives. Until the problems with facial recognition technology have been tackled, and regulations have been set, facial recognition technology in any shape or structure should be approached with caution.