Dataset for IDeAuth : A novel behavioral biometric-based implicit deauthentication scheme for smartphones
https://www.sciencedirect.com/science/article/abs/pii/S0167865522000770
IDeAuth Dataset contains boardcast data from 52 users. However, we discarded eleven testers due to insufficient data and used data of 41 users in our experiment. Among these 41 testers, 26 were males, 15 were females with 36 right-handed and 5 left-handed. Majority of testers were situated in Asia (19) and Europe (16) while per- forming the experiment, with 27 were between 20 to 30, 12 were between 30 to 40, and 2 were above 40.
6 motion sensors are used for features acquisition:
- Physical sensors - the accelerometer (Accel), the gravity sensor (Gravity), the gyroscope (Gyro), and the magnetometer (Magnet).
- Virtual/derieved sensors - the high-pass sensor (HPS) and the low-pass sensor (LPS).
A. From each sensor, 3 raw streams are acquired (X, Y and Z) and a magnitude is computed mathematically from it.
B. From each stream, 4 statstical features are determined that are Mean (μ), Standard Deviation (σ), Skewness (s), and Kurtosis (k) with 16
statistical features per sensor.
C. From 6 sensors, 16 x 9 = 96 sensors are obtained.
D. Overall, there are 96 statistical features per sample.
This can be used as header for csv file.
Accel_KurtX Accel_KurtY Accel_KurtZ Accel_KurtMag Accel_MeanMag Accel_MeanX Accel_MeanY Accel_MeanZ Accel_SkewMag Accel_SkewX Accel_SkewY Accel_SkewZ Accel_STDMag Accel_STDX Accel_STDY Accel_STDZ LPF_KurtX LPF_KurtY LPF_KurtZ LPF_KurtMag LPF_MeanMag LPF_MeanX LPF_MeanY LPF_MeanZ LPF_SkewMag LPF_SkewX LPF_SkewY LPF_SkewZ LPF_STDMag LPF_STDX LPF_STDY LPF_STDZ HPF_KurtX HPF_KurtY HPF_KurtZ HPF_KurtMag HPF_MeanMag HPF_MeanX HPF_MeanY HPF_MeanZ HPF_SkewMag HPF_SkewX HPF_SkewY HPF_SkewZ HPF_STDMag HPF_STDX HPF_STDY HPF_STDZ Magnet_KurtX Magnet_KurtY Magnet_KurtZ Magnet_KurtMag Magnet_MeanMag Magnet_MeanX Magnet_MeanY Magnet_MeanZ Magnet_SkewMag Magnet_SkewX Magnet_SkewY Magnet_SkewZ Magnet_STDMag Magnet_STDX Magnet_STDY Magnet_STDZ Gyro_KurtX Gyro_KurtY Gyro_KurtZ Gyro_KurtMag Gyro_MeanMag Gyro_MeanX Gyro_MeanY Gyro_MeanZ Gyro_SkewMag Gyro_SkewX Gyro_SkewY Gyro_SkewZ Gyro_STDMag Gyro_STDX Gyro_STDY Gyro_STDZ Gravity_KurtX Gravity_KurtY Gravity_KurtZ Gravity_KurtMag Gravity_MeanMag Gravity_MeanX Gravity_MeanY Gravity_MeanZ Gravity_SkewMag Gravity_SkewX Gravity_SkewY Gravity_SkewZ Gravity_STDMag Gravity_STDX Gravity_STDY Gravity_STDZ
Please cite our following papers to use the dataset.
@article{gupta2022ideauth,
title={IDeAuth: A novel behavioral biometric-based implicit deauthentication scheme for smartphones},
author={Gupta, Sandeep and Kumar, Rajesh and Kacimi, Mouna and Crispo, Bruno},
journal={Pattern Recognition Letters},
volume={157},
pages={8--15},
year={2022},
publisher={Elsevier}
}
@phdthesis{gupta2020next,
title={Next-generation user authentication schemes for iot applications},
author={Gupta, Sandeep},
year={2020},
school={PhD thesis, Ph. D. dissertation, University of Trento, Italy}
}