New features and improvements

  1. Updated VisionLabs SDK, added support for the Ada Lovelace GPU architecture.
  2. Updated FaceSDK VA: improved identification accuracy for the FR ALG2/ALG5 algorithms and face detection; added the liveness algorithms, including the detection of liveness by two frames and deepfake detection; improved the detection of mask and face attributes; added support for the NVIDIA Blackwell GPUs.
  3. You can now configure the detection of lifeless faces on the settings panel of the Face recognition server object 
  4. Added the verification parameters to the FIRSERVER|1|IDENTIFY_ACCEPT| and FIRSERVER|1|IDENTIFY_DECLINE| events from the face recognition server. The verification parameters include temperature temperature<>, alarm threshold temperature_alarm<>, face concealment evasion_int<>, emotion emotion_int<>, and a photo of the captured face imageBase64 for verification, taking into account the thermal imager and mask 
  5. Added the Enable face recognition upon request checkbox for the Face Detection object. The checkbox allows for capturing and recognizing faces upon request. If you set the checkbox, capture is enabled only upon an external request for a result from the camera and then disabled, which reduces the load and can be used for two-factor authentication 
  6. Added the request for getting the current number of faces in a frame of a specific camera to the REST API GET http://AxxonPSIM_user_login:AxxonPSIM_user_password@IP address:port/firserver/GetNumberFacesInFrame?server_id=server ID&capture_id=face detector ID 
  7. Improved the POST request http://AxxonPSIM_user_login:AxxonPSIM_user_password@IP address:port/firserver/FindPasserBy: the response now returns the camera ID from the configuration tree instead of the GUID, and the camera filter is processed correctly 
  8. Added the Number of requests in request history field to the Face recognition server object. Old requests are automatically deleted. The latest requests are saved 
  9. When you view search results in the Face search interface window, the track is displayed on the map with arrows between cameras. The track can be built for the selected result or the entire search page and include additional information (time and full name) 
  10. Added the ability to select GPU from the Select GPU drop-down list for the Face Detection object during recognition with VA and VL modules 
    You must set the Use GPU checkbox in the settings of the Face recognition server beforehand 
  11. Added the Use decoder NGP (GPU+CPU) checkbox to the settings panel of the Face recognition server object. If you set the checkbox, the GPU is used to decode video, reducing the load on the CPU 

Other

  1. Removed the Cognitec, KARS, STC, ISD integration server modules from Face PSIM 2.0.
  2. Updated and improved documentation.

Version features

Face PSIM 2.0 is compatible with Axxon PSIM 2.0. For the detailed information about the compatibility of product versions, see here.

You can download the current version of the Face PSIM 2.0 distribution package from the AxxonSoft official website or request it from technical support at support.axxonsoft.com.

When you update from an older version of Face PSIM, you don't need to uninstall the previous version, as the new version is installed over the old version. For details, see Installation.