Misconceptions about facial recognition technology
Ever since facial recognition technology became mainstream, there have been a lot of misconceptions and allegations against it. Honestly, some of them are true when the technology is in the wrong hands, like all other technology. However, arguing for a ban on technology stems from unawareness and misconceptions about the actual technology. Here are some of the most popular misconceptions about technology.
Hackers can track you with your facial ID.
It is a false conception that hackers can track you if they get your facial ID. Actually, it is less hackable than other types of IDs. Every identification system converts facial ID into a numerical code. This code won't be the same for other systems, since each facial recognition system has its own proprietary way of storing this data. So hackers can't really predict the face with the face ID. Whenever a man appears before a camera that recognises face, the system creates a facial ID and compares it with the facial IDs in the database.
Beside, everybody's face ID is public now on social media, which hackers can acquire easily. This is why we need a develop law that includes strict regulations around how and when you can add somebody to a faces to a system.
All face detection systems are the same.
Each system is different in how they create facial IDs and how they store the templates. Finding the right recognition system is very simple by determining how a face detecting system protects the rights and privacy of individuals. The accuracy and speed of the biometric system also vary. Some vendors earn money by selling data to third parties. Providers of good biometrics technology have established usage policies, built accurate algorithms with little bias, and adhered to privacy by design principles.
It can track anyone.
It is one of the popular controversies with face recognition systems. But, it can only track you or perform surveillance if your ID is in its database. If you have not opted to register your face with a company employee tracking system or you are on any system on a watchlist, then the system cannot track you.
Apart from this, tracking missing people with facial ID is one of the important applications of facial detection. This "watchlist" strategy can improve security at airports, schools, and other sites by identifying people of concern for that location right away. Using facial cameras to detect known threats is far more efficient and accurate than depending only on manual surveillance by security officers at each entry 24 hours a day, seven days a week.
Miss recognition may wrongly convict people.
Face recognition is one of the most useful tools for law enforcement agencies, but that doesn't mean it is the last word. Every matching and finding from the system is overlooked manually by humans. Even if an agency finds an exact match for a person in an exact scenario, the agencies have to verify him with other biometric characteristics. Usually, agencies collect ten fingerprints using slap scanners and the iris of culprits before allowing them into jails. So there is less possibility of misidentifying a person as a criminal.
Facial biometrics exhibits racial bias.
In the real world, the face detection camera can actually reduce facial bias. Police officers or agents are humans, so they can be biased about a person's color, ethnicity, and appearance, but not with facial APIs. A facial API will only show a person who is the real culprit.
To detect a person, the system may describe a person's color and ethnicity. Actually, it is a different case. How will you describe a person to someone else? With his physical appearance, right? That is only what the facial recognition system is also doing.