| | |
| | Updated anti-spoofing algorithm to improve the accuracy of liveness detection.
|
| | Updated anti-spoofing algorithm by adding dimensional detection of strong light. *Illumination detection and blurriness detection are turned off by default. Please turn it on manually.
|
| | Fixed several bugs to improve stability.
|
| | Updated anti-spoofing algorithm to improve the accuracy of liveness detection. Added a setting that requires the eyes to open in the returning images.
|
| | Updated anti-spoofing algorithm to improve the accuracy of liveness detection. Optimized the SDK internal processing logic to improve usability. Added the illumination condition detection function. Added code to change face frame in the sample.
|
| | Optimized the SDK internal processing logic to improve stability.
|
| | Optimized iOS error handling logic. Optimized the interface consistency.
|
| | Added re-detection interface. Updated anti-spoofing algorithm to improve the accuracy of liveness detection. Refactoring of the model loading related interfaces.
|
| | Reduced the SDK incremental volume. Updated the underlying algorithm to improve efficiency.
|
| | Reduced the installation package increment.
|
| | Optimized the project structure and integration experience. Fixed the problem for phones with the aspect ratio of over 2:1, which are easy to lose track. Adjusted adaptation of Sample UI on iPhone X.
|
| | Changed the name of the enumeration value of SilentLivefaceErrorType.
|
| | Updated model to improve the accuracy of blocked face detection. Added error prompts for eye, nose. and mouth blockage. Added return of request_id to facilitate record visits. Added the pass condition for silent liveness detection will return an error code when the setting is incorrect. Added a sample that stores the frame selection image locally in order to facilitate the recording of the issues of liveness detection.
|
| | Customized the judgment conditions of the face distance. Added camera orientation setting parameters for iOS input interface, supports external images/camera forwarding.
|
| | Updated the liveness detection model to increase the detection rate of spoofing attacks. Supported blocked face detection, such as by sunglasses or face masks. Released interface to configure path conditions for silent liveness detection to allow customization of the shortest detection time. Added state detection for faces that are completely within the screen to ensure the integrity of the face image collected by liveness detection. Improved error code consistency.
|
| | Updated the underlying algorithm library and model. Improved the passing rate for the correct person and the detection rate for spoofing.
|
| | Modified the delegate for successful liveness detection. Successful liveness detection will return the data_id, which is then made available to the cloud face comparison interface. Added an interface to return the face frame information of the result image to assist users to crop the face images. Optimized the detection of face distance when using the screen horizontally. Updated the model to improve the accuracy of face anti-spoofing detection.
|
| | Updated the underlying algorithm library. Updated the model to improve the accuracy of face anti-spoofing detection.
|
| | Added the online silent liveness detection SDK. Fixed the conflict when multiple SDKs are integrated together. Improved the consistency of processing methods for face status, distant face, close face, and track loss.
|
| | Provided a silent liveness detection solution to determine whether it is a correct person through facial hacking detection. Modified to collect high-quality face images during the liveness detection process. Modified return face image and binary encrypted data to be used with server interface. Added return face status information for each frame.
|
