3 Metrics

3.1 Location

The FRIF TE E1N application programming interface (API) can require that implementations include the finger position for a candidate on the candidate list. This is particularly useful when the probe template does not include a finger position. When the implementation is invoked in this way, search accuracy metrics are reported in terms of a locationregion or subject.

  • Region: The correct region of the correct subject was returned.
    • For probes sourced from a distal phalanx, the correct position 1–10 shall be returned.
    • For probes sourced from a palm or a non-distal phalanx, the most localized region shall be returned. Some palm regions may be interchangeable based on the exemplars provided (e.g., a palm probe’s source could reasonably be seen in a lower palm, hypothenar, and writer’s palm exemplar). Credit is given for Region in this case.
  • Subject: Any position from the correct subject is returned. This is designed to reward the situation where an implementation cannot ascertain the most localized region from the set of exemplars enrolled and may indicate segmentation error.

3.1.1 Notes

  • Multi-position probes are never requested to return a finger position.

3.2 Detection Error Tradeoff (DET)

The Detection Error Tradeoff (DET) plots in this document show the tradeoff between the False Positive Identification Rate (FPIR) and False Negative Identification Rate (FNIR) when searching probes against an enrollment database. For mated searches (used to compute FNIR), a single mated identity for each probe was present in the enrollment database. For non-mated searches (used to compute FPIR), there was no mate for the probe in the enrollment database.

3.2.1 Notes

  • The requested size of the candidate list was always 100 subjects.
  • The set of non-mated similarity scores come from the highest score when searching probes without a mate present in the enrollment database.
  • The set of mated similarity scores comes from searches of probes where the mate is present in the enrollment database and the algorithm successfully found the mate. The mate may appear at any rank in the candidate list.
  • Due to the quantity of searches, a sample of scores across the entire range of mated and non-mated scores were used to produce DET values.

3.3 Cumulative Match Characteristic (CMC)

The Cumulative Match Characteristic (CMC) plots in this document show the FNIR without respect for similarity score when searching probes against an enrollment database where a single mated identity for each probe was present. A description of the non-mated subject records represented in an enrollment database is documented in subsequent sections.

3.3.1 Notes

  • The metric hit rate is equivalent to 1 − miss rate, or 1 − FNIR. For example, an FNIR of 0.1 indicates a hit rate of 0.9 (i.e., 90%).