The case of wrong arrest occurred again. This "mistake" incident caused by facial recognition technology has also occurred in several regions of the world. In 2020, for example, the New Jersey Police, USA, announced that the man he arrested in connection with the theft at a hotel in Woodbridge in 2019 was proven not guilty.
The man was arrested after being identified using facial recognition, and police concluded that the technology was faulty after various investigations. Still in the US, the same incident also occurred in the same year in Detroit, where the local police announced that they had wrongly arrested after relying on facial recognition technology for a case of store theft in 2018.
So, what exactly is face recognition technology? How does it work and why can it lead the police to arrest people wrongly?
What is facial recognition?
Facial or face recognition is a technology that can be used to identify a person's face for various purposes. One of them is the process of investigating or searching for suspects in a case.
Based on information released by cybersecurity company Kaspersky, facial recognition technology will work gradually in order to accurately identify a person. First of all, a camera equipped with this technology will identify the targeted target's face.
After that, the face will be analyzed in detail using 2D or 3D imaging technology. There is some data taken here including face shape data, eyes, nose, and so on. Furthermore, the various facial data will be converted to digital data or information, and later will be adjusted with other facial data in the database or a system.
To determine the final result, the system will then use an algorithm to match the facial data that has just been taken with the facial data already stored in the system. The algorithm is used so that face recognition technology can adjust facial data in the database accurately and quickly.
"This high level of accuracy can be achieved when the camera takes pictures under ideal conditions, such as the correct position and lighting, as well as the face object being photographed without being obstructed (with various accessories)," CSIS wrote in a blog.
In addition, the position or angle of shooting to be identified can also be a determining factor for the level of accuracy of face recognition. When a photo of a face is taken in a free state, or says the target face is not ready to be photographed (laughing, expression doesn't fit, etc.) or is obscured by shadows, for example, the face recognition error rate can increase.
In addition, the age factor of the face owner can also reduce the level of facial recognition accuracy. This is because changes in facial contours will certainly make the system look for other facial data that matches the new facial contours. This difficulty occurs when the age of the photo of the face taken is a dozen years apart from the age of the photo of the face in the system. It should be noted that the level of accuracy or error in facial recognition may vary from vendor to vendor.
Currently, there are a number of vendors that provide facial recognition technology for various purposes. Some of these vendors include Kairos, Noldus, Affectiva, Sightcorp, Nviso, and many more.
Why is it still widely used? Although it can make people mistaken or misunderstood like some of the events above, facial recognition technology itself is still considered useful for various use cases. One of the closest to us is the use of the face unlock biometric feature that is on various smartphones today.
In addition to being used for the identification process of cell phone owners, this technology can also be used for law enforcement processes as mentioned above, as well as increasing efficiency in various user verification processes. At the airport, for example, many airport managers use this technology so that the check-in process or passenger verification can be done quickly without lining up.
This technology can also be used to search for missing people, verify customers at banks, verify patient data in hospitals, need attendance at the office, and can be used to improve user experience in various aspects.
However, this technology is also vulnerable to a number of issues such as privacy that is not maintained, prone to being stolen, storage that quickly runs out because the system has to store lots of facial photos, and is easily deceived and mistaken by many of these factors. Like two sides of a coin, a technology, no matter how sophisticated the technology, has the potential to be dangerous for those who use it, in addition to the various advantages and conveniences it offers.
