A recent study conducted by a team of researchers from Columbia University and the University of Buffalo has raised questions about the long-standing belief that every fingerprint is unique. The study, published in the journal Science Advances, utilized artificial intelligence technology to analyze a US government database of 60,000 fingerprints.
Contrary to popular belief, the researchers found that fingerprints from different fingers of the same individual exhibited strong similarities. Using an AI model called a deep contrastive network, the team was able to accurately determine when fingerprints belonged to the same person and when they didn’t, with a peak accuracy of 77%. This discovery challenges traditional fingerprint analysis methods that rely on minutiae, or specific features in ridge patterns.
According to Gabe Guo, the lead researcher on the project, the key to these similarities lies in the angles and curvatures at the center of the fingerprint. While traditional minutiae-based matching is effective for identification purposes, it lacks the ability to find correlations among fingerprints from the same person. The team’s findings suggest that the ridge orientation at the center of the fingerprint is a more reliable indicator of individuality.
Although the study acknowledges potential biases in the data and calls for further validation on a larger scale, Guo is confident that the discovery has significant implications for criminal investigations. By recognizing the similarities between fingerprints from different fingers, the research could potentially generate new leads in cold cases and prevent unnecessary investigations of innocent individuals.
However, some experts have expressed skepticism about the study’s findings. Christophe Champod, a professor of forensic science, argues that the correlation between fingerprint shapes across fingers has been known for years and has not been significantly utilized in practice. While acknowledging their attempt to match fingerprints from different fingers, he describes the study as “a tempest in a teacup.”
Despite the debate surrounding the study, there is no denying the potential impact of this new discovery. If further validated and integrated into forensic analysis, it could revolutionize the way fingerprints are used in criminal investigations and potentially improve the efficiency and accuracy of matching techniques.
The source of the article is from the blog regiozottegem.be