B2F (Biometric images of Fingerprints and Faces)

Citation Author(s):
Islem
JARRAYA
REGIM-Lab
Tarek M.
HAMDANI
REGIM-Lab
Adel M.
ALIMI
REGIML-Lab
Submitted by:
Islem Jarraya
Last updated:
Fri, 06/02/2023 - 06:48
DOI:
10.21227/3pwx-rb17
Data Format:
License:
829 Views
Categories:
Keywords:
0
0 ratings - Please login to submit your rating.

Abstract 

The B2F dataset (Biometric images of Fingerprints and Faces) has been prepared for face and fingerprint recognition, verification or classification.

The first subset (Fingerprint): This set of data presents the five finger feature vectors (of the left hand) for each person in a csv files.

The second subset (Face): This set of data presents feature vectors of face images in csv files. Feature vectors were extracted using the model (ResNet-50 + ArcFace). This set of face feature vectors represents:

- Feature vectors of face images taken on thirteen views (at a distance of 2 meters) (0°, 15°, 30°, 45°, 60°, 75°, 90°, 105°, 120°, 135°, 150 °, 165° and 180°),

- Feature vectors of face images taken with seven emotions (neutral, angry, happy disgusted, scared, surprised and sad)

- Feature vectors of four face images taken at three distances (two at a distance of two meters, one at one meter and one at a quarter of a meter) for each person.

Instructions: 

The "csv_testLabels" and "csv_test" files represent the labels and the feature vectors of the test part respectively.

The "csv_trainLabels" and "csv_train" files represent the labels and the feature vectors of the train part respectively.