Fingerprint Differential Morph Attack Detection

The aim of fingerprint differential morph attack detection is to assess whether a suspected fingerprint is a morph (also called a double-identity fingerprint) by comparing it to a bona fide fingerprint.

Benchmarks

Currently, two benchmarks are being used to evaluate the performance of fingerprint differential morph attack detection approaches:

• Fi-DMAD-1.0: A double-identity fingerprint benchmark created from the FVC2002 DB1-A database. It contains 800 real identity fingerprints from 100 fingers, 8 impressions per finger and three sets of 100 double-identity fingerprints produced using three different generation approaches (Fe, Im, and En) [1] [2] [3].
• Fi-DMAD-2.0: A double-identity fingerprint database created from the V300 Fingerpass database. It contains 8640 real identity fingerprints from 720 fingers, 12 impressions per finger and three sets of 700 double-identity fingerprints produced using three different generation approaches (Fe, Im, and En) [2] [3].

Performance Evaluation

For each algorithm, bona fide (evaluating bona fide fingerprints) and morph (evaluating double-identity fingerprints) attempts are performed to compute Bona fide Sample Classification Error Rate (BSCER) and Morphing Attack Classification Error Rate (MACER). As defined in [4] the BSCER is the percentage of bona fide samples incorrectly classified as morphed samples while the MACER is the proportion of morphing attack samples incorrectly classified as bona fide samples.

The following performance indicators are computed:

• EER (detection Equal-Error-Rate: the error rate for which both BSCER and MACER are identical)
• BSCER₁₀ (the lowest BSCER for MACER≤10%)
• BSCER₂₀ (the lowest BSCER for MACER≤5%)
• BSCER₁₀₀ (the lowest BSCER for MACER≤1%)

To calculate the above indicators, the following types of comparisons are conducted:

• Bona fide: each bona fide fingerprint is compared against the remaining bona fide impressions of the same finger. If fingerprint F_A is compared against F_B, the symmetric comparison is not executed to avoid correlation in the scores. The total number of bona fide comparisons is 2800 and 47520 for Fi-DMAD-1.0 and Fi-DMAD-1.0 benchmarks, respectively.
• Double-identity: each double-identity fingerprint is compared against all other bona fide impressions of both fingers involved in the generation process (excluding the two samples used in the generation process). The total number of double-identity comparisons for each double-identity set (i.e., Fe, Im, and En), is 1400 and 15840 for Fi-DMAD-1.0 and Fi-DMAD-1.0 benchmarks, respectively.

Results

The following table present the results obtained by the method proposed in [2] on the Fi-DMAD-1.0 benchmark.

The following table present the results obtained by the method proposed in [2] on the Fi-DMAD-2.0 benchmark.

Terms and Conditions

All publications and works that cite the information reported on this web page must reference sources [1], [2], [3] and [5].

Bibliography

[1] M. Ferrara, R. Cappelli and D. Maltoni, "On the Feasibility of Creating Double-Identity Fingerprints", IEEE Transactions on Information Forensics and Security, vol. 12, no. 4, pp. 892-900, April 2017.
[2] M. Ferrara, R. Cappelli and D. Maltoni, "Detecting Double-Identity Fingerprint Attacks", IEEE Transactions on Biometrics, Behavior, and Identity Science, vol. 5, no. 4, pp. 476-485, October 2023.
[3] M. Ferrara and A. Franco, "Deliverable 10.3 - Morphing techniques for secondary biometric characteristics", iMARS European Project, September 2023.
[4] ISO/IEC DIS 20059, "Methodologies to evaluate the resistance of biometric systems to morphing attacks", 2024.
[5] N. Gonçalves, "Deliverable 10.4 - Attack resolving solutions for finger, iris, and 3D face image manipulations", iMARS European Project, June 2024.