.png){kind=logo}
Facial recognition is a contentious technique. Although many of us have heard tales of AI mistaking humans for animals, that is only the beginning. We're just starting to understand how law enforcement uses artificial intelligence to spot criminals. However, when technology fails, the potential for disgrace is not all that we are faced with.
One American citizen was wrongly suspected of taking a pair of socks from a TJ Maxx store despite the fact that he was standing by his wife's side as she went into labor at the time of the incident. This case of mistaken identity is putting innocent people in jail.
Critics fear that incorrect identifications could result from facial recognition. What happens if a police department mistakenly associates someone shattering a store window during a disturbance with someone who wasn't even present? What is the likelihood that this might occur?
Depends, really. The best face identification program, according to tests by the National Institute of Standards and Technology, had an error rate of under 0.08 percent as of April 2020. In comparison to 2014, when the finest algorithm had an error rate of 4.1%, this represents a significant improvement.
According to a report by the Center for Strategic & International Studies (CSI) in 2020, accuracy is greater when identification techniques are used to identify people to clear, stable images, like a passport photo or mugshot. According to the report, when utilised in this manner, face recognition algorithms can achieve accuracy ratings of up to 99.97 percent on the Facial Recognition Vendor Test conducted by the National Institute of Standards and Technology.
However, accuracy rates are typically lower in the real world. According to the CSI report, the Facial Recognition Vendor Test discovered that when faces were compared against high-quality mugshots, the mistake rate for one algorithm was 0.1 percent, but when those faces were matched against photos of people taken in public, it climbed to 9.3 percent. Particularly when people weren't looking straight at the camera or were partially obscured by shadows or other objects, error rates increased.
Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp
_____________
Disclaimer: Analytics Insight does not provide financial advice or guidance. Also note that the cryptocurrencies mentioned/listed on the website could potentially be scams, i.e. designed to induce you to invest financial resources that may be lost forever and not be recoverable once investments are made. You are responsible for conducting your own research (DYOR) before making any investments. Read more here.
