Releases: 3DiVi/face-sdk
v3.25.1
v3.25.0
What's New
-
Faster extraction of biometric templates and reduced model weights. A lightweight version is now available for each of the up-to-date recognition methods. While minimizing accuracy degradation, lightweight models provide increase in inference speed ( ~40-50% for mobile CPU and ~20-25% for desktop CPU) an ~60% reduction in model weights used in template extraction.
-
Improved accuracy of Liveness Detection methods. We continue to improve the accuracy of our Liveness Detection algorithms. In this release you'll see a 40% increase in attack detection accuracy (decrease APCER@BPCER=0.05 on all attack types).
-
Redesigned TemplateIndex - functional and user-friendly. We redesigned the module responsible for storing the template database and 1:N search. Now you can:
- Dynamically (without recreating TemplateIndex) add and remove templates from the database.
- Name each template with an arbitrary alphanumeric ID (previously you had to store this information separately).
- In case of highly loaded applications, an asynchronous version of the new TemplateIndex can be used, which will allow simultaneous addition/removal of elements and 1:N searches.
-
Testing methods for evaluating SDK quality. This is a detailed guide, including the main tasks addressed by Face SDK, the quality metrics used and a set of scripts for testing. With this guide you'll be able to quickly test the core technologies of Face SDK for your specific use case and see how well it meets your needs.
-
Support for accelerated inference on Rockchip NPU. Support for NPU inference from the Rockchip manufacturer has been added for the Face detection and Face Template Extractor modules. Using the new inference will significantly speed up calculations even on the company's low performance devices.
Bug Fixes and Improvements
-
Filtering detection results by bbox size. Ability to filter detection bboxes by width and height. Ideal for filtering out small faces in the background.
-
Combining multiple detectors into a cascade to improve accuracy in complex cases. When you need to detect faces in the same entrypoint on different domain data, such as selfies, ACS camera photos and ID card photos, it can be difficult to ensure high recognition quality. Combine different face detectors into a cascade so you don't miss a single face!
-
C# Face SDK API update with support for .NET 8. Due to the expiration of the LTS for .NET 6, we have migrated our C# Face SDK API to .NET 8.
-
Fixed a bug in the SetBytes method of the Context class in Java and the Kotlin API that led to segfault on 32-bit devices.
v3.24.2
Bug Fixes and Improvements
- Fixed an issue in the Flutter API that caused memory leaks on iOS.
- Fixed an issue in the Flutter API that caused incorrect image conversion to RGB format on some devices.
- Fixed an issue that periodically triggered license blocking after an emergency device reboot.
- Fixed an issue in the QUALITY_ASSESSMENT_ESTIMATOR processing block with the second version of the assessment modification, which occasionally led to a segfault.
v3.24.1
Bug Fixes and Improvements
- Fixed an error in the MATCHER_MODULE processing block that led to incorrect search results.
- Resolved an issue with angle detection during strong head turns for the tddfa fitter.
- Corrected a bug in the EYE_OPENNESS_ESTIMATOR that occasionally caused a segfault.
- Improved the display of results in processing_block_demo.
- Updated the Flutter version; Face SDK now supports versions 3.24.x.
v3.24.0
What's New
- Added the GLASSES_ESTIMATOR processing block to assess the presence of glasses on a face.
- Introduced a new version of the QUALITY_ASSESSMENT_ESTIMATOR processing block with modified assessment capabilities. This version estimates facial image quality according to the international ICAO standard.
- Added a new ssyv_light modification for the FACE_DETECTOR processing block, offering high accuracy in access control and remote identification scenarios, and is also suitable for mobile devices.
- Added support for processing blocks in the Swift API.
- Introduced new versions for the LIVENESS_ESTIMATOR processing block with 2d and 2d_light modifications, which reduce recognition errors during attacks by half.
- Added 2d_ensemble and 2d_ensemble_light modifications to the LIVENESS_ESTIMATOR processing block. These new modifications perform better against 2D and 3D mask attacks.
Bug Fixes and Improvements
- Reduced the size of the libfacerec.so library for the Android platform.
- Improved angle detection accuracy for the FDA fitter.
- Fixed errors in the Flutter API that caused segfaults on 32-bit devices.
- Fixed an issue in the FACE_FITTER processing block that caused an exception when the objects array was missing in the input context.
v3.23.0
What's New
- Added EYE_OPENNESS_ESTIMATOR processing block.
- A modification has been added to the FACE_FITTER processing block for calculating anthropometric points from the fda set.
- Added functions for saving and loading Context-container into json files.
- Added support for Kotlin language for Android.
Bug Fixes and Improvements
- FACE_FITTER processing blocks additionally output head rotation angles.
Face SDK v3.22.1
Bug Fixes and Improvements
- Fixed bugs that caused the application to crash in the Python API.
- Fixed bugs that caused the application to crash in the Java API on Windows.
- Fixed the behavior of the uld detector, where reducing the min_size parameter led to a decrease in face detection quality.
- Fixed a bug for the FACE_TEMPLATE_EXTRACTOR processing block, where the intra_op_num_threads configuration parameter did not affect the operation of the processing block.
Face SDK v3.22.0
What's New
- It's now possible to utilize the Processing Blocks API in the Java API.
- Flutter API now includes classes with isolates.
- Added auxiliary functions for image reading, conversion, and input Context-container formation.
Bug Fixes and Improvements
- Minor fixes for the Python API.
Face SDK v3.21.0
What's new
- Added version 2 of the 2d processing block for Liveness estimation.
- New 2d_light modification added to the Liveness estimation processing block with faster and lighter algorithms.
Bug Fixes and Improvements
- Improved error messaging with Face SDK version output included.
Face SDK v3.20.1
Bug Fixes and Improvements
- Fixed app_id generation for Android 11+.
- Resolved a licensing error when running multiple Face SDK processes from different users.
- Fixed improper functioning of VideoWorker in Flutter.