In recent times, biometrics based on brain (electroencephalogram) and heart (electrocardiogram) signals have emerged.[23][24] The research group at University of Kent led by Ramaswamy Palaniappan has shown that people have certain distinct brain and heart patterns that are specific for each individual. The advantage of such ‘futuristic’ technology is that it is more fraud resistant compared to conventional biometrics like fingerprints. However, such technology is generally more cumbersome and still has issues such as lower accuracy and poor reproducibility over time.
Perhaps I expressed myself poorly. What I meant was that if I were attempting to disguise myself for whatever reason, I can change my gait to walk as “Not Oak” fairly easily, unlike my fingerprints which are not something I can readily change.
On an unrelated point, when I went on flight status in the Air Force, they took my footprints. I asked why, and was told that if I was involved in a crash, they might be able to identify my corpse by footprint, since my feet would be in heavy leather flightboots…and might be all they were able to recover.
I used to take standardized tests at a local Pearson Vue testing center (about 2010-2011) and they used this technology to make sure the test-taker was who he said he was. The first time I went there I had to bring a boatload of identifying documents and have both my palms scanned. After the first time they required only a driver’s license and a palm scan to match the ones on record. Evidently something went wrong with the scan of the right palm the first time because after that it never matched. The left palm always matched and they let me take the tests. Either that or I cut off the left hand of the real bibliophage and attached it to my own left wrist. How would they know?
Perhaps it’s more about the difficulty/(potential) impossibility of matching someone else’s gait? Like, you can change your gait to walk as “Not Oak”, but you can’t walk exactly as “Elm” or “Birch” (though probably pretty close, especially for short periods of time).
I recently completed a Coursera online course, where they asked for a webcam head shot before the class started, and when doing the quizzes. They also asked for you to type a short phrase that they said they would use to verify the same person typed the phrase each time. I can’t believe a typing pattern would be unique, but it was good enough for them. Anyone have stats on typing patterns?
Yes, but the *Mission Impossible *movie scene was not about changing your gait - of course anyone can change their gait temporarily - it was about trying to imitate someone else’s gait in order to pass oneself off as someone else.
So, you cannot imitate my gait well enough, and I cannot imitate your gait well enough.
As a side note to the walking gait thing, some years ago I was able to identify a brand-new coworker as the younger sister of another coworker, entirely based on how she moved. The two sisters looked nothing alike, had dramatically different personalities, I had not yet heard the younger woman’s last name, and I had as yet not seen the two of them together. But on her second day on the job, after seeing her perform some tasks (she had the same job as her sister), I realized that her movements were uncannily familiar. I asked, and she confirmed that, yes, she was the other woman’s sister.
During a big awards show a comic looked out over a roomful of celebrities, with their perfect teeth. He said if the auditorium were suddenly destroyed by fire, most of the people in the room couldn’t be identified by their teeth.
While the OP focuses on something not being fake-able, for ID purposes it also has to be reliable for that person. Each and every time a person gets checked, they will pass.
A walk is horrible for this. An ankle strain, big toe stub, different pair of shoes (sneakers vs. high heels, for example) would all ruin it.
If you write the software to be generous enough to handle these situations, the ease of fakery increases significantly.