The False Accept Rate or FAR (similar to the False Match Rate or FMR) is a measurement of the accuracy of biometric data being recorded by a given system. By its very nature biometric data is organic in origin, and organic data is subject to considerable variance between both individuals and sessions.
With both interface and security-based biometrics, FAR is referred to as the accuracy of the data being checked. A high FAR indicates that there is a wide margin for error in the data being checked, such that whilst it is possible to pick up all variations an individual might use – differences in their gait for example – opening the net that far may incorrectly identify other individuals as being that person.
On the other hand, narrowing the margin for error too much, will effectively lock out an authorised person who maybe has a sore ankle that day, and is walking slightly differently.
With interface biometrics, the situation is a little different. Here biometrics are usually neural codes or EMG readings of muscle potential. A high false accept rate here means that there is a high potential that the neural code will be misread, and an action other than what was intended, triggered. As an extreme example, the person meant to raise their hand with the beer glass in it, but the system falsely interpreted that as another command instead, opening the hand so the beer glass fell to the floor and smashed.
See Also: Biometric, Biometrics, Neural Code, False Match Rate, FMR, False Reject Rate, False Non-Match Rate, EMG
Below, we offer a selection of links from our resource databases
which may match this term.
Entries for FAR:
in our database matching the Term FAR:
Results by page
News containing the Term FAR:
Results by page