Researchers at US-based Columbia University have challenged the longstanding notion that a person's each fingerprint is unique, using an AI tool.
The study, published in journal Science Advances, and conducted by a team led by Columbia Engineering undergraduate senior Gabe Guo, used a US government database of more than 60,000 fingerprints and fed them into an AI tool known as deep contrastive network.
"Over time, the AI system, which the team designed by modifying a state-of-the-art framework, got better at telling when seemingly unique fingerprints belonged to the same person and when they didn’t. The accuracy for a single pair reached 77%.
"When multiple pairs were presented, the accuracy shot significantly higher, potentially increasing current forensic efficiency by more than tenfold," the researchers said.
To Professor Hod Lipson, one of the authors of the study, it was clear that the model wasn't using traditional markers that forensics have been using for decades. "It seems like it is using something like the curvature and the angle of the swirls in the centre."
Lipson said both he and Guo were surprised by the outcome. "We were very sceptical ... we had to check and double check."
The study, co-authored with Professor Wenyao Xu of the University of Buffalo, has challenged the widely accepted belief that each fingerprint is unique.
The team faced rejection from multiple journals before finally getting their work accepted by Science Advances.
Guo recounted the initial pushback from the forensics community, highlighting the lack of a forensics background in his team. However, perseverance and a commitment to data-driven analysis led the team to present compelling evidence that fingerprints may not be as unique as conventionally believed.
Despite the groundbreaking nature of their findings, the team acknowledged potential biases in their data and emphasised the need of a larger and more diverse fingerprint database for the AI system to be viable in actual forensics.
Guo expressed confidence that their discovery could significantly impact criminal investigations, potentially generating new leads for cold cases and preventing unnecessary investigations of innocent individuals.
However, not everyone in the forensic community is entirely convinced with the findings.
Christophe Champod, a professor of forensic science at University of Lausanne in Switzerland, suggested that the study may have oversold its findings, arguing that the correlation between fingerprints from different fingers has been known for years.
Simon Cole, a professor in criminology at University of California, Irvine, questioned the practical utility of the study, asserting that the unproven claim of fingerprint uniqueness remains intact.
Guo defended the work, asserting his team is the first to systematically quantify and utilise the similarities between fingerprints from different fingers of the same person.
The team open-sourced the AI code for verification, emphasising the importance of its study not just in forensics but as a testament to the potential of AI in uncovering patterns hidden in plain sight.
"Even more exciting is the fact that an undergraduate student, with no background in forensics whatsoever, can use AI to successfully challenge a widely held belief of an entire field. We are about to experience an explosion of AI-led scientific discovery by non-experts, and the expert community, including academia, needs to get ready," said Lipson.
Comments