PARIS – “Have you ever wondered if every fingerprint is truly unique?”
That’s the question that sparked Gabe Guo’s curiosity during the COVID-19 lockdowns. As a freshman at Columbia University, he found himself engrossed in a conversation with a professor that would ultimately shape the next three years of his life.
Now a senior in Columbia’s Department of Computer Science, Guo led a groundbreaking study that challenges the long-held belief that every fingerprint is distinct. Teaming up with Professor Wenyao Xu of the University of Buffalo, their research, published in Science Advances, has stirred up controversy in the forensics community.
Uncovering Surprising Results
The study utilized an artificial intelligence model called a deep contrastive network to analyze a U.S. government database of 60,000 fingerprints. The AI system found that fingerprints from different fingers of the same person shared strong similarities, debunking the notion of absolute uniqueness.
Guo explained, “We found a rigorous explanation for why this is the case: the angles and curvatures at the center of the fingerprint.”
While the study acknowledges potential biases in the data, Guo is confident that this discovery can revolutionize criminal investigations. “This can help generate new leads for cold cases and prevent innocent people from being unnecessarily investigated,” he said.
Debating the Findings
However, not everyone is convinced. Forensic science professor Christophe Champod believes that the study has merely rediscovered something known for years. On the other hand, Simon Cole, a professor in criminology, law, and society, acknowledges the study’s interest but questions its practical utility.
Despite the skepticism, Guo and his team stand by their results and have open-sourced the AI code for others to verify. Guo believes that this study is just the beginning of AI’s potential to uncover hidden truths that have been overlooked by humans for centuries.
Doubtful, we’ve relied on fingerprints for so long! #NotConvinced