To All QVR Face Users
We are pleased to announce the official release of QVR Face 2.3.0.
As AI image recognition continues to expand, system performance, compatibility with various video sources, and integration capabilities have become crucial factors in building a smart surveillance environment.
In this update, we have expanded NVIDIA GPU hardware acceleration support for QVR Face, now supporting NVIDIA RTX 50 series GPUs. We have also added features such as real-time fisheye correction and face recognition for specified regions, multi-group profile management, real-time detection variables for Event URL, and event forwarding to Nx VMS. With these enhancements, you can now build a more efficient, flexible, and highly integrated facial recognition system with greater ease.
Next-gen Hardware Performance: Now Supports NVIDIA RTX 50 Series GPUs
QVR Face 2.3.0 introduces support for the NVIDIA RTX 50 series GPUs (requires QVR AI Library 1.1.0 or later). Utilizing hardware acceleration can boost AI video analysis performance. In environments where multiple video streams are processed simultaneously or under heavy face recognition workloads, GPU acceleration effectively distributes processing tasks, enhancing recognition efficiency and deployment flexibility. Accelerate your QVR Face now!
Expanded Use of Fisheye Cameras: Real-Time Fisheye Correction and Face Recognition for Specified Regions
Fisheye cameras can monitor wide areas, but image distortion near the edges may impact the accuracy of face recognition. QVR Face 2.3.0 addresses this by supporting real-time fisheye correction and face recognition for specified regions. Users can select specific areas for correction and execute facial recognition, reducing blind spots with fisheye cameras and extending the coverage of a single device for entrances, open spaces, and wide-area surveillance scenarios.
More Flexible Data Classification: Add Profiles to Multiple Groups
In real-world management environments, the same person may belong to multiple departments, permission groups, or usage categories. QVR Face 2.3.0 now allows “profiles” to be added to multiple groups. This gives administrators greater flexibility in classifying and managing personnel data, making it easier to maintain large facial databases. For example, the same employee can be categorized under “Headquarters Employees,” “R&D Department,” and “VIP Access List” at the same time, simplifying later searches, filtering, and event utilization.
Enhanced Event Integration: Event URLs Now Support Real-Time Detection Variables and Custom Event Forwarding to Nx Witness
In QVR Face 2.3.0, we’ve strengthened event integration by enabling Event URLs to include real-time detection variables. Recognition results, event details, and related variables can now be sent to third-party platforms, event dashboards, access control systems, and other automated processes, making event handling and display even more flexible.
This update also further enhances integration with Nx Witness. Face recognition events and detection variables from QVR Face can be sent to your existing Nx Witness system, allowing tailored event notifications and surveillance flows for environments already using Nx Witness, and further improving interoperability and event management efficiency. (For event integration, your Nx Witness VMS version must support the REST Generic Event API.)
Improvements to Usability and System Stability
Along with new features, QVR Face 2.3.0 introduces several enhancements to usability and system performance:
- Supports the creation of face recognition tasks that connect to QVR Surveillance with two-factor authentication enabled.
- Added notifications for manual uploads of similar images, improving profile creation efficiency and preventing duplicate entries.
- Added sorting options for unregistered faces to improve data management convenience.
- Adjusted Chinese language search mechanisms for improved search operations.
Key Updates in This Version
| Feature Category | Update Details |
|---|---|
| Hardware Compatibility | Support for NVIDIA RTX 50 series GPUs |
| Fisheye Video Usage | Real-time fisheye correction & face recognition by area |
| Profile Management | Add profiles to multiple groups; filter/sort unregistered faces |
| QVR Client Integration | Create profile photos from detection snapshots |
| Real-Time System Notifications | Event URLs support real-time detection variables |
| Advanced Brand Integration | Send facial recognition events to Nx Witness |
| System Stability | Optimized sync mechanism, storage protection, Chinese search, and bug fixes |
Download QVR Face 2.3.0 Now!
Experience faster GPU acceleration, more flexible fisheye video recognition, more convenient data management, and more comprehensive inter-system event integration now.
For more details, please see the release notes: App Release Notes | QNAP
Thank you for your continued support of QNAP and the QVR product series. We are committed to further developing smarter, more stable, and highly integrated AI surveillance solutions.