Biometric authentication, identification and detection method and device for mobile terminal and equipment

1. A method for attack detection in biometric authentication, implemented by a mobile terminal device, the method comprising:

obtaining a risk level of a user corresponding to the mobile terminal device;

determining a time length for collecting sensor data of the mobile terminal device based on the obtained risk level of the user, wherein the sensor data of the mobile terminal device comprises movement information of the mobile terminal device, and wherein the time length for collecting sensor data of the mobile terminal device is increased with an increased risk level of the user;

performing the biometric authentication to authenticate the user of the mobile terminal device based on biometric information of the user;

collecting the sensor data of the mobile terminal device for the determined time length, while the mobile terminal device performs the biometric authentication, wherein the sensor data comprises acceleration sensor data of the mobile terminal device indicating accelerations of the mobile terminal device in x, y, and z axes of a three-dimensional space, and/or gyroscope data of the mobile terminal device indicating angular velocities in the x, y, and z axes of the three-dimensional space; and

inputting the sensor data and the risk level of the user into an attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using the risk level of the user and sensor training data obtained based on sensor data of the mobile terminal device collected when the mobile terminal device performed biometric authentication historically.

2. The method according to claim 1 , wherein the attack determination model includes a determination threshold, and the determination threshold is set based on the risk level of the user corresponding to the mobile terminal device.

3. The method according to claim 1 , further comprising:

obtaining information of a model of the mobile terminal device; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the sensor data and the information of the model of the mobile terminal device into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using sensor training data corresponding to the model of the mobile terminal device.

4. The method according to claim 1 , wherein before inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication, the method further comprises:

pre-processing the sensor data by performing at least one of the following operations on the sensor data: filtering operation, boundary detection operation, or length normalization operation, to obtain pre-processed sensor data; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the pre-processed sensor data into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using pre-processed sensor training data which is obtained by pre-processing the sensor training data.

5. The method according to claim 1 , wherein before inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication, the method further comprises:

performing feature extraction on the sensor data to obtain the sensor data with features extracted; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the sensor data with features extracted into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using sensor training data with features extracted which is obtained by performing feature extraction on the sensor training data.

6. The method according to claim 5 , wherein the acceleration sensor data of the mobile terminal device is measured by an acceleration sensor of the mobile terminal device, and the gyroscope data is measured by a gyroscope of the mobile terminal device; and

wherein the sensor data with features extracted comprises:

an average, a standard deviation, a median, a maximum, a minimum, and/or an integral of measurements of the acceleration sensor of the mobile terminal device in the x, y, and z axes of the three-dimensional space; and/or

an average, a standard deviation, a median, a maximum, a minimum, and/or an integral of measurements of the gyroscope of the mobile terminal device in the x, y, and z axes of the three- dimensional space.

7. The method according to claim 5 , wherein the sensor data with features extracted also comprises a time length of performing the biometric authentication.

8. The method according to claim 1 , further comprising:

adjusting a security authentication level of the biometric authentication based on an output result of the attack determination model, wherein the higher a probability of attack occurrence is indicated by the output result, the higher the security authentication level of the biometric authentication becomes after the adjustment.

9. An apparatus for attack detection in biometric authentication, comprising one or more processors and one or more non-transitory computer-readable memories coupled to the one or more processors and configured with instructions executable by the one or more processors to cause the apparatus to perform operations comprising:

obtaining a risk level of a user corresponding to a mobile terminal device;

determining a time length for collecting sensor data of the mobile terminal device based on the obtained risk level of the user, wherein the sensor data of the mobile terminal device comprises movement information of the mobile terminal device, and wherein the time length for collecting sensor data of the mobile terminal device is increased with an increased risk level of the user;

performing the biometric authentication to authenticate the user of the mobile terminal device based on biometric information of the user;

collecting the sensor data of the mobile terminal device for the determined time length while the mobile terminal device performs the biometric authentication, wherein the sensor data comprises acceleration sensor data of the mobile terminal device indicating accelerations of the mobile terminal device in x, y, and z axes of a three-dimensional space, and/or gyroscope data of the mobile terminal device indicating angular velocities in the x, y, and z axes of the three-dimensional space; and

inputting the sensor data and the risk level of the user into an attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using the risk level of the user and sensor training data obtained based on sensor data of the mobile terminal device collected when the mobile terminal device performed biometric authentication historically.

10. The apparatus according to claim 9 , wherein the attack determination model includes a determination threshold, and the determination threshold is set based on the risk level of the user corresponding to the mobile terminal device.

11. The apparatus according to claim 9 , wherein the operations further comprise: obtaining information of a model of the mobile terminal device; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the sensor data and the information of the model of the mobile terminal device into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using sensor training data corresponding to the model of the mobile terminal device.

12. The apparatus according to claim 9 , wherein before inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication, the operations further comprise:

pre-processing the sensor data by performing at least one of the following operations on the sensor data: filtering operation, boundary detection operation, or length normalization operation, to obtain pre-processed sensor data; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the pre-processed sensor data into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using pre-processed sensor training data which is obtained by pre-processing the sensor training data.

13. The apparatus according to claim 9 , wherein before inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication, the operations further comprise:

performing feature extraction on the sensor data to obtain the sensor data with features extracted; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the sensor data with features extracted into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using sensor training data with features extracted which is obtained by performing feature extraction on the sensor training data.

14. The apparatus according to claim 13 , wherein the acceleration sensor data of the mobile terminal device is measured by an acceleration sensor of the mobile terminal device, and the gyroscope data is measured by a gyroscope of the mobile terminal device; and

wherein the sensor data with features extracted comprises:

an average, a standard deviation, a median, a maximum, a minimum, and/or an integral of measurements of the acceleration sensor of the mobile terminal device in the x, y, and z axes of the three-dimensional space; and/or

an average, a standard deviation, a median, a maximum, a minimum, and/or an integral of measurements of the gyroscope of the mobile terminal device in the x, y, and z axes of the three- dimensional space.

15. The apparatus according to claim 13 , wherein the sensor data with features extracted also comprises a time length of performing the biometric authentication.

16. The apparatus according to claim 9 , wherein the operations further comprise:

adjusting a security authentication level of the biometric authentication based on an output result of the attack determination model, wherein the higher a probability of attack occurrence is indicated by the output result, the higher the security authentication level of the biometric authentication becomes after the adjustment.

17. A non-transitory computer-readable storage medium configured with instructions executable by one or more processors to cause the one or more processors to perform operations comprising:

obtaining a risk level of a user corresponding to a mobile terminal device;

determining a time length for collecting sensor data of the mobile terminal device based on the obtained risk level of the user, wherein the sensor data of the mobile terminal device comprises movement information of the mobile terminal device, and wherein the time length for collecting sensor data of the mobile terminal device is increased with an increased risk level of the user;

performing the biometric authentication to authenticate the user of the mobile terminal device based on biometric information of the user;

collecting the sensor data of the mobile terminal device for the determined time length while the mobile terminal device performs the biometric authentication, wherein the sensor data comprises acceleration sensor data of the mobile terminal device indicating accelerations of the mobile terminal device in x, y, and z axes of a three-dimensional space, and/or gyroscope data of the mobile terminal device indicating angular velocities in the x, y, and z axes of the three-dimensional space; and

inputting the sensor data and the risk level of the user into an attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using the risk level of the user and sensor training data obtained based on sensor data of the mobile terminal device collected when the mobile terminal device performed biometric authentication historically.

18. The non-transitory computer-readable storage medium according to claim 17 , wherein the attack determination model includes a determination threshold, and the determination threshold is set based on the risk level of the user corresponding to the mobile terminal device.

19. The non-transitory computer-readable storage medium according to claim 17 , wherein the operations further comprise: obtaining information of a model of the mobile terminal device; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the sensor data and the information of the model of the mobile terminal device into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using sensor training data corresponding to the model of the mobile terminal device.

20. The non-transitory computer-readable storage medium according to claim 17 , wherein before inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication, the operations further comprise:

pre-processing the sensor data by performing at least one of the following operations on the sensor data: filtering operation, boundary detection operation, or length normalization operation, to obtain pre-processed sensor data; and

wherein inputting the sensor data into an attack determination model to determine whether an attack occurs in the biometric authentication comprises:

inputting the pre-processed sensor data into the attack determination model to determine whether an attack occurs in the biometric authentication, wherein the attack determination model is trained by using pre-processed sensor training data which is obtained by pre-processing the sensor training data.