Frequently Asked Questions

Q1. Introduction to Face Verification SDK authorization method

  1. Online activation file authorization method

    Only for the first activation, device needs to be connected to the network, after that the generated activation code can be saved and read locally every time.

    1)Direct activation (network connection required for the first activation): use the authorization file provided.

    2)Step-by-step activation (for devices that cannot be connected to network):

Provides UUID acquisition tool, online activation authorization file and website for step-by-step activation. Clients can upload the generated uuid and the downloaded authorization file to "Step-by-step Activation Website" to generate an activation code. After passing the activation code into the correct SDK interface, the SDK is ready for normal use.

Note: When using online activation file authorization method, pay attention to whether the hardware information of the device has been changed, including the CPU, all wired and wireless network cards, Bluetooth and 4G network cards. Changes in hardware may cause changes in uuid, resulting in repeated consumption of the number of authorization activations.

  1. Chip authorization method

1)Offline authorization (unable to count the number of activated devices) + chip.

2)Online activation file authorization (able to count the number of activated devices) + chip. >

Note: Chip authorization method is to bind with the chip, which will not be affected by changes in the hardware information of the devices (including CPU, all wired and wireless network cards, Bluetooth and 4G network cards).

Q2. When using online activation file authorization method to run Face Verification SDK demo on the same device, why is the number of authorization activations consumed repeatedly?

For a given device, if the same license is used, and the uuid of the corresponding device has not changed, then no matter how many activations are performed on this device, the consumption of the number of activations is only one. The following are two common reasons for repeatedly consuming authorization activations:

1)Device uuid change:

UUID is an unique identifier of a device. The “Online activation file authorization method” will automatically obtain the UUID of the device for activation. However, UUID will be affected by changes in device hardware information (including CPU, all wired and wireless network cards, Bluetooth and 4G network cards), therefore the change of device hardware information will cause the change of its uuid, resulting in repeated consumption of the number of authorization activations.

2)A different authorization file is used (open the authorization file xxx.lic to check the unique identifier of the authorization License SN, which can be used to distinguish different authorizations).

Q3. Error code -13

Error code -13 indicates that the license is invalid or the license file is not loaded. Possible reasons are: >

1)License capability & SDK capability mismatch: For example, a license with only monocular liveness detection capability is being used in a binocular liveness detection SDK;

2) Interface for loading the license has not been called. For details, please refer to the provided SDK demo.

3)Wrong uuid acquisition tool is used for online activation file authorization method (step-by-step activation), such as using the Face Verification SDK V1.2.1 UUID acquisition tool for V2.0.1 and later versions, resulting in the the generation of a wrong activation code. The uuid acquisition tool can be found in SDK download package.

4) Issue with the license. Open the license file and send the license serial number in plain text format to our support to receive technical support.

Q4. Error code -15

Error code -15 indicates that the license file or SDK has expired. There are three reasons: >

1)Open xxx.lic to check whether the license has expired. If so, a new license need to applied.

2)Check whether the system time of the device is correct;

3) If the first two steps are passed, delete the original activation code generated in the folder of the device or restore the device to factory settings, and run the sample again afterwards. The previously installed and tested sample may have been using an old license (which has been expired); the activation file generated online stays on the device and causes an error.

Q5. Error code -29

Error code -29 indicates that the online activation code is incorrect. Normally due to the damage of the activation file or the change of the device hardware information. Check whether the hardware information has been changed recently first. Possible causes are:

1)The same device was in normal use, and -29 was reported suddenly: >

The possible reason is that the hardware information of the client's device (not necessarily the replacement of the hardware) has changed, including the CPU, all wired and wireless network cards, Bluetooth and 4G network cards, which caused changes to the unique hardware identifier UUID, which is read by the SDK. This error occurs when UUID is mismatched.

【Note: For details, please refer to Q1. Introduction to Face Verification SDK authorization method】

2) New devices deployed by client report -29:

The possible reason is that the activation file already exists on client's device, which was copied from another device, therefore, the UUID bound to this activation file is from the other device.

Q6. What is the reason that when using the activation code generated by V1.2 and earlier versions in V2.0.x and later versions, error -29 was reported?

The online activation file authorization method of the Face Verification SDK can be performed by direct activation or step-by-step activation:

1)For direct activation, activation code generated by the Face Verification SDK 1.2 and previous versions cannot be used in the Face Verification 2.0.x version;

2)For step-by-step activation, activation code generated by the Face Verification SDK 1.2 and previous versions cannot be used in the Face Verification 2.0.x version. The uuid acquisition tool has been updated for 2.0 and later versions.

Q7. Error code -33?

Error code -33 indicates that the encrypted chip verification failed. There are two possible reasons:

1)The license and chip are not from the same batch;

2)32-bit libraries are running on 64-bit devices.

Q8. When running SDK sample, error message "Invalid Camera ID" is displayed.

This error is caused by the camera driver. The SDK part is not related to the camera. Check if there are other samples that can turn on two cameras at the same time. Move the method the camera is turned on into our SDK sample.

Q9. Error: toExternalResult:sdk failed, result -196608

Error code -196608 indicates a networking error.

Check whether the network connection is weak and test whether the device has the proxy mode turned on. When using the online activation file authorization method, the device needs to be connected to the network when performing the activation for the first time, and the network connection is no longer needed after activation.

Q10. Error JNI DETECTED ERROR IN APPLICATION: thread Thread[32,tid=5956,Native,Thread=0xd19cdc00,peer=0x12d4d820,"pool-3-thread-1"] using JNIEnv from thread Thread[32,tid=5956,Native,Thread*=0xd19cdc00,peer=0x12d4d820,"pool-3-thread-1"]

This error is caused by a call error to jni, resulting in thread disorder. In response to this problem, need to check whether the same method is called in different threads, causing thread processing disorder.

Note: The current version SDK handle does not support multi-threaded calls, if multiple threads need to be called for multiple times, multiple handles should be created and each thread should have a separate handle.

Q11. What are the supported architectures of the Face Verification SDK on Android platform? How to configure to make it support armeabi architecture?

Our SDK aar includes two types of architectures: arm64-v8a & armeabi-v7a; the SDK only supports these two architectures by default.

If you need armeabi at the same time, you can copy the so file in armeabi-v7a from the SDK aar package to armeabi in different projects.

Note: Architectures of x86, x86_64, mips and mips64 cannot be supported.

Q12. What are the configuration requirements for Face Verification SDK?

1)Android

CPU: Qualcomm Snapdragon 800, MediaTek MT6795, Rockchip RX3399 or above (armv7 or aarch64 architecture, supports NEON technology).

Available memory: 1GB or more.

Available storage space: 1GB or more.

System version: Android 5.0 and above.

2)Windows >

CPU: Intel Core 4th Generation (Haswell) i5 quad-core 2.8GHz and above.

Available memory: 4GB or more.

Available storage space: 2GB or more.

System version: Windows 7 and above.

3)Linux

CPU: Intel Core 4th Generation (Haswell) i5 quad-core 2.8GHz and above.

Available memory: 4GB or more.

Available storage space: 2GB or more.

System version: Ubuntu 14.04 and above / CentOS 7 and above / Redhat 7 and above.

4)ARM Linux

A CPU performance of not less than rk3399 is recommended.

General Linux system, general gcc compiler (gcc 4.8 or 4.9), armv7.

Available memory: 1GB or more.

Available storage space: 1GB or more.

Q13. Does the Face Verification SDK support the integration of monocular & binocular functions into one project at the same time?

1)Face Verification SDK 1.2.1 and earlier

Does not support the integration of monocular and binocular functions into the same project at the same time.

2)Face verification SDK 2.0.x and later

Supports the integration of monocular and binocular functions into the same project at the same time.

Q14. Does distorted image affect the result of detection and recognition?

Distorted images have an impact on detection. It is not recommended to use cameras such as "ball cameras" that obtain distorted images.

Q15. How many faces can be tracked by binocular liveness detection?

Binocular liveness detection can only track one face at a time (there are two cameras in the binocular, the images captured by each camera are different. To ensure the consistency of the face tracked by these two cameras, all binocular liveness detection currently can track maximum one face).

Q16. Can different versions of SDK or operating system platforms share the same feature values?

No, that is not possible.

Q17. Return error code -30 ST_E_ONLINE_ACTIVATE_CONNECT_FAIL.

This is an indication that the online verification connection failed. Please check whether the device can be connected to the network.

Last updated