Login
Public area
Home
Background
Benchmarks
Register
Published Results
Statistics
Participant area
Login
Upload
Pending Algorithms
Tested Algorithms
Download
Past editions
FVC2006
FVC2004
FVC2002
FVC2000
Result of algorithm MagFace 1 on DMAD-SOTAMD_P&S-1.0
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
4.83% (4.83% - 4.84%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
2.37%(F:2.37% )
Efficiency indicators
Avg Detection Time
Avg Detection Time (BF)
Avg Detection Time (M)
0.4 s
0.4 s
0.4 s
Memory indicators
Max Detection Memory
3.4 GB
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Summary of Subset Accuracy Indicators
Subset
Bona Fide
Morphing
EER
BSCER
10
BSCER
20
BSCER
100
Morphing Factor=0.3
10960
18530
3.67%
0%
0%
0%
Morphing Factor=0.5
10960
37000
5.31% (5.31% - 5.32%)
0%
0%
0%
Female
4600
24880
5.03% (5.02% - 5.04%)
0%
0%
0%
Male
6360
30650
4.66% (4.65% - 4.66%)
0%
0%
0%
African Ethnicity
1600
13700
5.59% (5.56% - 5.62%)
0%
0%
0%
East Asian Ethnicity
530
5050
0.56% (0.55% - 0.57%)
0%
0%
0%
European/American Ethnicity
7600
34410
5.08%
0%
0%
0%
Indian/Asian Ethnicity
660
4620
2.50% (2.42% - 2.58%)
0%
0%
0%
Middle Eastern Ethnicity
570
5320
2.35% (2.28% - 2.42%)
0%
0%
0%
Age: 18-35
6360
36290
2.68% (2.67% - 2.68%)
0%
0%
0%
Age: 36-55
3480
24320
5.98% (5.98% - 5.99%)
0%
0%
0%
Age: 56-75
1120
8230
8.24% (8.21% - 8.26%)
0%
0%
0%
Trait: Freckels
800
8840
3.18% (3.13% - 3.24%)
0%
0%
0%
Trait: Moles
1260
13370
10.01% (10.00% - 10.03%)
0%
0%
0%
No Specific Trait
7590
47900
3.21% (3.20% - 3.21%)
0%
0%
0%
Hochshule Darmstadt Images
5000
13500
3.02% (3.01% - 3.02%)
0%
0%
0%
Norwegian University of Science and Technology Images
2960
18030
1.65% (1.65% - 1.66%)
0%
0%
0%
University of Twente Images
3000
24000
8.77%
0%
0%
0%
Compressed Images
8960
49830
5.13% (5.13% - 5.14%)
0%
0%
0%
Uncompressed Images
2000
5700
2.91% (2.90% - 2.93%)
0%
0%
0%
Automatic Post-processing
10960
21790
6.30%
0%
0%
0%
Manual Post-processing
10960
33740
3.97% (3.97% - 3.98%)
0%
0%
0%
Morphing Algorithm: C01
10960
7500
5.47%
0%
0%
0%
Morphing Algorithm: C02
10960
7500
2.83%
0%
0%
0%
Morphing Algorithm: C03
10960
18950
4.30%
0%
0%
0%
Morphing Algorithm: C05
10960
14080
4.75% (4.74% - 4.75%)
0%
0%
0%
Morphing Algorithm: C06
10960
7500
9.04% (9.03% - 9.04%)
0%
0%
0%
Manual Post-process: PM01
10960
3000
1.11% (1.10% - 1.12%)
0%
0%
0%
Manual Post-process: PM02
10960
7500
2.83%
0%
0%
0%
Manual Post-process: PM03
10960
6160
2.87% (2.86% - 2.87%)
0%
0%
0%
Manual Post-process: PM05
10960
9580
3.65%
0%
0%
0%
Manual Post-process: PM06
10960
7500
9.04% (9.03% - 9.04%)
0%
0%
0%
Low Quality Morphed Images
10960
20560
4.16%
0%
0%
0%
High Quality Morphed Images
10960
34970
5.11%
0%
0%
0%
Overall
10960
55530
4.83% (4.83% - 4.84%)
0%
0%
0%
Subset: Morphing Factor=0.3
Attempts
Bona Fide
Morphing
10960
18530
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
3.67%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
2.38%(F:2.38% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Morphing Factor=0.5
Attempts
Bona Fide
Morphing
10960
37000
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.31% (5.31% - 5.32%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
2.37%(F:2.37% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Female
Attempts
Bona Fide
Morphing
4600
24880
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.03% (5.02% - 5.04%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.17%(F:1.17% )
1.56%(F:1.56% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Male
Attempts
Bona Fide
Morphing
6360
30650
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
4.66% (4.65% - 4.66%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.89%(F:1.89% )
3.03%(F:3.03% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: African Ethnicity
Attempts
Bona Fide
Morphing
1600
13700
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.59% (5.56% - 5.62%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
5.25%(F:5.25% )
4.73%(F:4.73% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: East Asian Ethnicity
Attempts
Bona Fide
Morphing
530
5050
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
0.56% (0.55% - 0.57%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0%
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: European/American Ethnicity
Attempts
Bona Fide
Morphing
7600
34410
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.08%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.18%(F:1.18% )
1.95%(F:1.95% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Indian/Asian Ethnicity
Attempts
Bona Fide
Morphing
660
4620
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.50% (2.42% - 2.58%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0%
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Middle Eastern Ethnicity
Attempts
Bona Fide
Morphing
570
5320
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.35% (2.28% - 2.42%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0%
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Age: 18-35
Attempts
Bona Fide
Morphing
6360
36290
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.68% (2.67% - 2.68%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0.57%(F:0.57% )
0.86%(F:0.86% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Age: 36-55
Attempts
Bona Fide
Morphing
3480
24320
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.98% (5.98% - 5.99%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.38%(F:1.38% )
1.38%(F:1.38% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Age: 56-75
Attempts
Bona Fide
Morphing
1120
8230
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
8.24% (8.21% - 8.26%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
8.04%(F:8.04% )
8.14%(F:8.14% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Trait: Freckels
Attempts
Bona Fide
Morphing
800
8840
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
3.18% (3.13% - 3.24%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
4.50%(F:4.50% )
3.19%(F:3.19% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Trait: Moles
Attempts
Bona Fide
Morphing
1260
13370
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
10.01% (10.00% - 10.03%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.90%(F:1.90% )
4.85%(F:4.85% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: No Specific Trait
Attempts
Bona Fide
Morphing
7590
47900
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
3.21% (3.20% - 3.21%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.50%(F:1.50% )
2.16%(F:2.16% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Hochshule Darmstadt Images
Description: Images acquired at Hochshule Darmstadt location.
Attempts
Bona Fide
Morphing
5000
13500
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
3.02% (3.01% - 3.02%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0%
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Norwegian University of Science and Technology Images
Description: Images acquired at Norwegian University of Science and Technology location.
Attempts
Bona Fide
Morphing
2960
18030
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
1.65% (1.65% - 1.66%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0%
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: University of Twente Images
Description: Images acquired at University of Twente location.
Attempts
Bona Fide
Morphing
3000
24000
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
8.77%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
5.80%(F:5.80% )
5.49%(F:5.49% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Compressed Images
Description: Images stored using a lossy image format (JPG, JP2000).
Attempts
Bona Fide
Morphing
8960
49830
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.13% (5.13% - 5.14%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.94%(F:1.94% )
2.64%(F:2.64% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Uncompressed Images
Description: Images stored using a lossless image format.
Attempts
Bona Fide
Morphing
2000
5700
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.91% (2.90% - 2.93%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
0%
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Automatic Post-processing
Description: Morphed images obtained without manual post-processing.
Attempts
Bona Fide
Morphing
10960
21790
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
6.30%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
3.59%(F:3.59% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Manual Post-processing
Description: Morphed images obtained applying manual post-processing.
Attempts
Bona Fide
Morphing
10960
33740
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
3.97% (3.97% - 3.98%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
1.59%(F:1.59% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Morphing Algorithm: C01
Description: Hochshule Darmstadt method based on FaceMorpher.
Attempts
Bona Fide
Morphing
10960
7500
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.47%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
3.48%(F:3.48% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Morphing Algorithm: C02
Description: Hochshule Darmstadt method based on FaceFusion.
Attempts
Bona Fide
Morphing
10960
7500
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.83%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Morphing Algorithm: C03
Description: Norwegian University of Science and Technology method based on FaceMorph.
Attempts
Bona Fide
Morphing
10960
18950
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
4.30%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
1.38%(F:1.38% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Morphing Algorithm: C05
Description: University of Bologna method based on triangulation with DLIB landmarks.
Attempts
Bona Fide
Morphing
10960
14080
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
4.75% (4.74% - 4.75%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
2.56%(F:2.56% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Morphing Algorithm: C06
Description: University of Twente method based on triangulation with STASTM-landmarks.
Attempts
Bona Fide
Morphing
10960
7500
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
9.04% (9.03% - 9.04%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
5.80%(F:5.80% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Manual Post-process: PM01
Description: Hochshule Darmstadt manual post-processing using GIMP.
Attempts
Bona Fide
Morphing
10960
3000
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
1.11% (1.10% - 1.12%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Manual Post-process: PM02
Description: Hochshule Darmstadt manual post-processing using GIMP.
Attempts
Bona Fide
Morphing
10960
7500
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.83%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Manual Post-process: PM03
Description: Norwegian University of Science and Technology manual post-processing using Photoshop.
Attempts
Bona Fide
Morphing
10960
6160
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
2.87% (2.86% - 2.87%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
0%
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Manual Post-process: PM05
Description: University of Bologna manual post-processing using GIMP.
Attempts
Bona Fide
Morphing
10960
9580
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
3.65%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
1.04%(F:1.04% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Manual Post-process: PM06
Description: University of Twente manual post-processing using GIMP.
Attempts
Bona Fide
Morphing
10960
7500
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
9.04% (9.03% - 9.04%)
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
5.80%(F:5.80% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: Low Quality Morphed Images
Description: Morphed images rejected at face verification stage by at least one face recognition SDK between Neurotechnology and Cognitec.
Attempts
Bona Fide
Morphing
10960
20560
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
4.16%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
3.36%(F:3.36% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Subset: High Quality Morphed Images
Description: Morphed images accepted at face verification stage by both Neurotechnology and Cognitec face recognition SDKs.
Attempts
Bona Fide
Morphing
10960
34970
Accuracy indicators
EER
BSCER
10
BSCER
20
BSCER
100
5.11%
0%
0%
0%
Detection failures
REJ
NBFRA
REJ
NMRA
1.59%(F:1.59% )
1.79%(F:1.79% )
Graphs
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Score distributions
MACER(t) and BSCER(t) graphs
DET graph
Description of algorithm MagFace 1
Algorithm type:
NIST dynamic library
Docker image:
nist_frvt_ubuntu-20.04.3
A - Brief description
This algorithm is based on the deep neural network approchach proposed by Scherhag et al. But in contrast to the base algorithm, MagFace embeddings are used instead of the ArcFace embeddings.
B - References
Ulrich Scherhag, Christian Rathgeb, Johannes Merkle, and Christoph Busch. Deep face representations for differential morphing attack detection, 2020 Qiang Meng, Shichao Zhao, Zhida Huang and Feng Zhou. MagFace: A universal representation for face recognition and quality assessment, CVPR, 2021
For information or suggestions:
fvc@unibo.it
Copyright © 2025 Biometric System Laboratory