What is Facial Recognition Technology?
This is the technology which allows a machine to identify an individual from an image or a video. Can you imagine that someone with this technology can just use a random picture you took from somewhere to find out who you are, and get as many details about you as possible.
This innovation has been around for quite a long time, however, its utilization has turned out to be progressively perceptible, and available, in the previous couple of years as it currently powers innovative solutions, for example, individual photograph applications and secondary authentication for cell phones. To comprehend these rising capacities, let’s first talk about how facial recognition functions.
Amazon Facial Recognition
Amazon has called their facial recognition applications Amazon Rekognition. Facial investigation abilities, for example, those accessible in Amazon Rekognition, enable clients to comprehend where faces exist in a picture or video, just as what characteristics those appearances have.
For instance, Amazon Rekognition can dissect properties, for example, eyes open or shut, temperament, hair shading, just as the visual geometry of a face.
These recognized traits become progressively valuable for clients that need to compose or seek through a large number of pictures in seconds utilizing metadata labels (e.g., glad, glasses, age extend) or to distinguish an individual (i.e., facial acknowledgement utilizing either a source picture or a special identifier).
How does Amazon’s Rekognition works?
Faces are matched dependent on their visual geometry, including the connection between the eyes, nose, temples, mouth, and other facial highlights. At the point when pictures are dissected by Amazon Rekognition, there is a diagram around the face, called a bounding box, which decides the main piece of the picture recognition considers in its examination. The investigation at that point produces object documentation numbers for the picture that show the “area” for the real components of the face. At the point when clients are running a face look, the innovation is contrasting this information from the source picture to every one of the pictures it seeks. From that point, the administration relegates each face in the picture a comparability score. This methodology guarantees that Amazon Rekognition has no data about the personality of an individual, just the probability that one face is a potential counterpart for another.
How accurate is Facial Recognition
Lots of people have called into question the authenticity of the abilities of these technologies. In comparison to human’s ability to recognize faces, how accurate will these technologies be?
Some people have the conception that humans can better match faces than any form of technology. But a report by the National Institute for Standards and Technology (NIST) released the outcome of a study on facial recognition technologies that predate that of Amazon.
According to the study, even the older technologies could outperform people’s ability to recognise faces.
More so, since the technology is based on artificial learning, it can only get better with time. This is because this technology will continuously learn and improve upon itself. Inadvertently, false positives or false negatives may be encountered but these will be reduced with time.
Is Rekognition a violation of human right?
Imagine innocent people being surcharged with crimes they know nothing about?
Imagine someone looking forward to harming easily identifying you without anyone’s help?
Can there be many other negative applications to this technology?
These set of questions are critical when it comes to appraising the impact of such a technology in this modern era. Many people have raised lots of concern about the potential misuse of this technology but this has not scared Amazon off to put a halt to their production of this technology.
The Securities and Exchange Commission (SEC) ruled that Amazon puts to vote a ban placed by the government on the sales of their technology.
Shareholders of Amazon, therefore, voted on the ban and only 2% of them voted for the motion. The proposed ban was not to stop sales altogether, yet rather to confine contracts with police and other government offices until a full and autonomous request had surveyed the potential dangers.
Amazon keeps up that Rekognition helps keep the public safe, yet different executions over the world have not been exhibited to be completely dependable.
The greatest concern of Amazon is on the potential benefits of this technology which they think may far outweigh the risks it possesses. For instance, this technology can help us to easily identify missing people or accurately identify suspects.
What do you think about the uses of this technology? Leave your comments below