Sign up for CNN’s Wonder Theory science newsletter. Explore the universe with news about fascinating discoveries, scientific developments and more.
“Do you think every fingerprint is actually unique?”
It’s a question a professor asked Gabe Guo during an informal conversation as he sat at home during Covid-19 lockdowns, waiting to start his freshman year at Columbia University. “Little did I know that this conversation would become the basis for the focus of my life for the next three years,” Guo said.
Guo, now a student at Columbia’s computer science department, led a team researching the topic, with University at Buffalo professor Wenyao Xu as one of his co-authors. The paper, published this week in the journal Science Advances, appears to turn a long-accepted truth about fingerprints on its head: Guo and his colleagues argue that not all of them are unique.
In fact, journals rejected the work several times before the team appealed and ultimately had it accepted by Science Advances. “There was a lot of resistance initially from the forensic community,” recalls Guo, who had no background in forensics before the investigation.
“For the first two versions of our article, they said it is a well-known fact that no two fingerprints are the same. “I think this really helped improve our study because we put more and more data into it (increasing its accuracy) until eventually the evidence was irrefutable,” he said.
A new look at old prints
To achieve surprising results, the team used an artificial intelligence model called a deep contrastive network, which is often used for tasks such as facial recognition. The researchers added their own twist and then fed it to a US government database of 60,000 fingerprints in pairs that were sometimes from the same person (but from different fingers) and sometimes from different people.
As it worked, the AI-based system discovered that fingerprints from different fingers of the same person showed strong similarities and was therefore able to determine with single-pair accuracy when the fingerprints were from the same person and when they were not. peaking at 77% – seemingly disproving the idea that every fingerprint is ‘unique’.
“We found a rigorous explanation for why this is the case: the corners and curves in the center of the fingerprint,” Guo said.
Over hundreds of years of forensic analysis, he added, people have looked at various features called “minutiae,” the branches and endpoints in fingerprint ridges that are used as the traditional markers for fingerprint identification. “They are great for matching fingerprints, but not reliable for finding correlations between fingerprints of the same person,” says Guo. “And that’s the insight we had.”
The authors said they are aware of possible biases in the data. While they believe the AI system works in much the same way across genders and races, the study says more careful validation through the analysis of a larger and broader database of fingerprints is required for the system to be useful in actual forensic investigations.
However, Guo said he is confident the discovery can improve criminal investigations.
“The most immediate application is that it can help generate new leads for cold cases, where the fingerprints left at the crime scene come from fingers other than those in the file,” he said. “But on the other hand, this will not only help catch more criminals. This actually helps innocent people who may no longer need to be examined unnecessarily. And I think that’s a victory for society.”
‘A storm in a teacup’?
The use of deep learning techniques on fingerprint images is an interesting topic, according to Christophe Champod, professor of forensic sciences at the School of Criminal Justice at the University of Lausanne in Switzerland. However, Champod, who was not involved in the research, said he does not believe the work has uncovered anything new.
“Their argument that these shapes are somewhat correlated between the fingers has been known since the very beginning of fingerprinting, when it was done manually, and it has been documented for years,” he said. “I think they oversold their newspaper due to lack of knowledge. I’m glad they’ve rediscovered something familiar, but essentially it’s a storm in a teacup.”
In response, Guo said that no one had ever quantified or used the similarities between fingerprints from different fingers of the same person as systematically as in the new study.
“We are the first to explicitly point out that the similarity is due to the orientation of the edge at the center of the fingerprint,” says Guo. “Moreover, we are the first to attempt to match fingerprints from different fingers of the same person, at least with an automated system.”
Simon Cole, a professor in the department of criminology, law and society at the University of California, Irvine, agreed that the article is interesting but said its practical usefulness is overstated. Cole was also not involved in the investigation.
“We weren’t ‘wrong’ about fingerprints,” he said of forensic experts. “The unproven but intuitively true claim that no two fingerprints are ‘exactly alike’ is not refuted by the observation that fingerprints look alike. It has always been known that fingerprints from different people, as well as from the same person, look alike.”
According to the newspaper, the system could be useful at crime scenes where the fingerprints found come from fingers other than those in the police file, but Cole said this could only happen in rare cases because when prints are taken, all ten fingers and often palms are routinely recorded. “It is not clear to me when they think that the police will register only some, but not all, of an individual’s fingerprints,” he said.
The team behind the study says it is confident in the results and has made the AI code open source for others to check, a decision that both Champod and Cole praised. But Guo said the importance of the investigation goes beyond fingerprints.
“This isn’t just about forensics, it’s about AI. People have been looking at fingerprints since we existed, but no one ever noticed this similarity until we had our AI analyze it. That just speaks to the power of AI to automatically recognize and extract relevant features,” he said.
“I think this study is just the first domino in a huge chain of these things. We will see people use AI to discover things that were literally hiding from our eyes, like our fingers.”
For more CNN news and newsletters, create an account at CNN.com