Study Finds Fingerprints Less Unique Than Previously Thought, Raising Questions About Identification Methods
-
Study uses AI to analyze 60,000 fingerprint pairs, finds fingerprints are not all unique as previously thought, with accuracy up to 77% in determining if prints are from same person
-
Researchers focused on ridge orientation at the center of prints rather than traditional minutiae features used in fingerprint analysis
-
Immediate application could help generate leads in cold cases where prints at crime scene are different fingers than prints on file
-
Study rebutted by some fingerprint experts who say similarities between prints, including from same person, have been long known
-
Authors argue their AI analysis quantified intra-person similarities more rigorously than before and has uncovered new insights, though practical applications questioned by critics