3DiVi is proud to announce the release of the latest version of our Face Recognition Image API, now upgraded to 2.0.0. This new version introduces significant advancements in biometric technology, streamlining processes and enhancing accuracy for a wide range of applications.
New Template-Extractor Service
A standout feature in this release is the "template-extractor" module, designed to efficiently extract a face template using a set of precise anthropometric points. This service operates without the need for initial face detection, significantly increasing the speed of template extraction compared to our previous "face-detector-template-extractor."
Enhanced Specificity and Pipeline Integration
The new service is adept at building templates for specific faces, a marked improvement over the previous version which automatically generated templates for every detected face in an image. Users can now seamlessly integrate service pipelines, such as "face-detector-face-fitter -> template-extractor," streamlining the process for more complex tasks.
Liveness Estimation with Greater Accuracy
The addition of the "liveness-estimator" module represents another major step forward. Although this service operates at a slightly slower pace, it offers remarkably improved accuracy in liveness estimation, outperforming the earlier face-detector-liveness-estimator.
Face SDK Updated to Version 3.19.0
Alongside these new services, the Face SDK has been updated to version 3.19.0. This update includes presets of face detection settings tailored for specific scenarios such as safe city initiatives, remote identification via selfies, access control systems, and more. This feature allows users to employ a face detector configured for specific tasks without the need to manually select optimal parameters. (Read more about Face SDK 3.19.0)
Enhanced Documentation and Error Reporting
The updated documentation now includes a comprehensive list of potential errors returned by the services. This addition is crucial for users to quickly identify the causes of failures and implement effective solutions.
API Versioning and Standardized Data Pattern
With the introduction of API versioning, we have also rolled out a new version "v2" for all services. This update brings a standardized data pattern across all Face SDK, Image API, and Omni Platform products. The standardized format simplifies cross-product queries and enhances the translation of processing results from one product to another.
About 3DiVi Image API
The 3DiVi Image API is a comprehensive suite of software tools specifically designed for advanced biometric analysis. Key features include detection of faces and silhouettes, face verification, and estimation of facial attributes from various image formats. It offers capabilities like age, gender, and emotion estimation, face mask detection, liveness checks, image quality assessment, and extraction of biometric templates. Additionally, it supports a standardized sample format across Face SDK, Image API, and OMNI Platform for easier cross-product integration. The API is tailored for diverse applications including safe city projects, remote selfie identification, and both cooperative and non-cooperative recognition in access control systems.