-: FOLLOW US :- @theinsaneapp
It is widely accepted that fingerprints from different fingers of the same person are unique and unmatchable.
-: FOLLOW US :- @theinsaneapp
Led by Columbia Engineering undergraduate Gabe Guo, an AI system was developed to challenge this presumption.
-: FOLLOW US :- @theinsaneapp
Guo used a U.S. government database of 60,000 fingerprints, employing a deep contrastive network for analysis.
-: FOLLOW US :- @theinsaneapp
The AI system achieved 77% accuracy in determining if seemingly unique fingerprints belonged to the same person.
-: FOLLOW US :- @theinsaneapp
With multiple pairs, the accuracy significantly increased, potentially improving forensic efficiency tenfold.
-: FOLLOW US :- @theinsaneapp
The project involved collaboration between Columbia Engineering and University at Buffalo, SUNY labs.
-: FOLLOW US :- @theinsaneapp
A well-established forensics journal initially rejected the findings, citing the widely held belief in fingerprint uniqueness.
-: FOLLOW US :- @theinsaneapp
The team persisted, gathering more data, and appealed the rejection, emphasizing the importance of the discovery.
The findings were finally published in Science Advances, challenging and surprising the forensics community.