In general, accuracy is defined as a state of being correct or precise. Accuracy boosts the quality of the system. In biometrics, accuracy is measured by taking pre-defined samples and it is processed further to measure the accuracy. It is important to note that this works in a closed environment to prevent interference with any real-time values.

The important key performance metrics are

  1. False Match Rate (FMR) 
  2. False Non Match Rate (FNMR)
  3. Failure to Enroll (FTE)

Analyzing of all the three metrics is necessary to access the performance of a specific technology


The first metric in the biometric system is FMR(False Match Rate). In this, one user template is wrongly identified or incorrectly judged to be a match for some other user. 

Consider two users, user1 & user2.

In a verification system, the user1 approaches the biometric system and enters the username or ID of user2, and presents the biometric data. It is successfully matched. Now, user1 gets access to operate user2 resources( physical or logical). There is no initial identity claim in the identification system. So from the system’s database, he/she(user1) is identified as user2. 

FMR is also called False Acceptance Rate. False match rates can be reduced by adjusting the thresholds that adjust the level of correlation. 


As biometrics play a major role in today’s world, it is highly important to make solutions to provide high security, since FMR is the most critical accuracy metric. FMR must always be balanced with FNMR and FTE rates.


In general, FNMR is the probability that a user’s template will not get matched with their own templates. In the verification system, the user1 approaches a biometric system, entering the username or ID, and presenting the biometric data but it fails to match. In the identification system, the user1 approaches the biometric system in which he/she had previously registered but not being located/identified in the database at that time. FNMR is referred to as False Rejection Rate(FRR)

FNMR occurs because of less correlation between the user’s verification and enrollment templates. This occurs due to the external factors

  • Change in biometric data of the user- It may be temporary or permanent. Temporary changes include sweating, injury, etc., and permanent changes include aging and scar formation
  • Change in the presentation of data- The way the user is giving the input such as presenting data from different angles and wearing spectacles.
  • Environmental changes-The changes here are lighting, background noise, and temperature.


When users are falsely nonmatched and denied access it leads to an increased burden. It also makes users frustrated. A disproportionate focus on FMR reduction can lead to unacceptable high FNMR rates.


A system’s FTE rate represents the probability that a given user will be unable to enroll in the biometric system. It is inaccessible due to technical issues. But it can be rectified by giving prior instructions to the staff. Physiological and behavioral biometrics are subjected to FTE. A system’s FTE rate is dependent on system design, training, and ergonomics and does not only depend on biometric data.


Improvement in system design and enrollment process can reduce a system’s FTE rate. The impact of FTE differs for individual and institutional users. For individuals, if it is unable to enroll then it will be frustrating. But in institutions, if it is unable to enroll then it becomes a customer service issue. So in these environments, high FTE rates are linked directly to increased security risks and system costs.

Click here to know more about applications of nanofibres!

Click here to know more about the story of Hyundai!