A study conducted at Columbia University questions the common belief in the uniqueness of fingerprints.
Using an artificial intelligence tool that examined 60 thousand fingerprints, researchers tested whether prints from different fingers belonged to the same person.
The artificial intelligence in question can detect which person’s fingerprints belong to with a 75 – 90 percent accuracy rate.
This research challenges the widespread belief that fingerprints are completely unique. It shows that the artificial intelligence tool can match prints from different fingers with a very high accuracy rate.
However, researchers are still skeptical about the situation. Because they believe that the artificial intelligence tool analyzes fingerprints in a different way than traditional methods.
By focusing specifically on the orientation of the folds in the middle of the finger, it is thought to take a different approach from traditional methods regarding the way individual folds, called minutiae , terminate and bifurcate.
Research; It reveals that it could have potential impacts on biometrics and forensic science.
For example, an unidentified thumb print at crime scene A and an unidentified index finger print at crime scene B cannot currently be forensically linked to the same person, but an AI tool can make that connection.
It is stated that artificial intelligence tools are generally trained on large amounts of data and more fingerprints will be needed to develop this technology.
In addition, while it is pointed out that all fingerprints used to develop the model are complete and of high quality, it is emphasized that incomplete or poor-quality prints can often be encountered in the real world.
A woman living in England claims that her grandchildren have an interesting talent.
Allegedly, the twin grandchildren, who looked alike when they were born, became different over time. But what’s surprising is the grandchildren’s ability to bypass phones’ facial recognition, not just with their fingerprints.
That means grandchildren can bypass security measures with both their fingerprints and facial recognition technology. This strengthens claims that fingerprints may not be unique.
