DetectorPack 3.12.1 Release Notes

Compatibility

This version of DetectorPack is compatible with Axxon One 2.0 and later versions. 

What's new:

1. Supported Nvidia GPUs with Ada Lovelace micro-architecture for:

   • Neurotracker
   • Neurocounter
   • Pose Detection
   • Equipment detection (PPE)
   • Fire Detection
   • Smoke Detection
   • Object Presence Detection
   • Barcode Detection
   • Water Level Detection
   • Stopped object detector
   • Privacy Masking
   • Person-based privacy masking
   • Face Detection VA
   • Crowd Estimation VA
   • Abandoned Objects Detection VI
   • Abandoned Objects Detection (Street) VI
   • License plate recognition RR
   • License Plate Recognition VT

2. Added Stopped object detector

3. Added Object Tracker VL Object Tracker VL

4. Added Glow detector

5. Added Vehicle Recognition XR Windows OS.
6. Added new countries for License Plate Recognition IV
7. Added new countries for the License plate recognition RR 

8. Added new countries for the License Plate Recognition VT 

9. It is now possible to hide the license plate number completely or partially using the GDPR Policy feature.

10. It is now possible to create events based on unrecognized license plates for detection tools:

    • Configuring the License Plate Recognition RR

    • Configuring License Plate Recognition RR—Parking

Optimization

1. Optimized the Nvidia GPU utilization so it can be used more efficiently, which increased the performance of the detection tools:

   • Neurotracker
   • Fire Detection
   • Object Presence Detection
   • Smoke Detection
   • Neurocounter

2. Added new embedded neural network for the Person (top-down view) for detection tools:

   • Neurotracker
   • Stopped object detector
   • Neurocounter

3. Added the possibility of using object classes in standard detection sub-tools

4. Updated the integration of the Abandoned Objects Detection VI and Abandoned Objects Detection (Street) VI—Windows version is now available.

5. Added demo license for the detection tools:

   • Face detector VL
   • PPE detection VL
   • Object Tracker VL

Update recommendations

1. For Linux OS: if a detector pack 3.7.4 version and earlier that includes all analytics tools is installed (not divided into plugins), then before updating, it is necessary to remove the installed package without using purge (complete package deletion) by running the following command:
   

   sudo dpkg --remove --force-depends axxon-detector-pack

   and then install the necessary plugins.

2. For Face masking to work during export, it is necessary to have the Neuro Pack installed.

3. By default, the GDPR Policy feature for detection tools created before the update will be disabled. To enable it, it is necessary to delete the field values ​​and then re-enter '-' or specify the required filter.

Update limitations

1. When you update the Face Detection TV, Face Detection VA or Face Detection VL (see Installing DetectorPack add-ons), the results of the previously recognized persons will become unavailable. If you want to keep the previously recognized persons, do not update the analytics package.
2. To start and correctly run the basic Face Detection and Face recognition VA on GPU, you must first create a cache
3. Replaced Face Detection TV with Face Detection VA. Discontinued the support for Face Detection TV. Face Detection TV and Face Detection VA are incompatible.
4. Discontinued the support for the Intel HDDL (Mustang-v-100) devices.
5. The mechanisms for launching neural network models have changed, that is why some of the custom models (trained specifically for the object, see Data collection requirements for neural network training) may not work correctly. If you encounter any difficulties, please contact technical support.
6. In this version of DetectorPack, you cannot filter objects by color when using sub-tools (child detection tools).
7. When working with Fire Detection and Smoke Detection in scanning mode (see Configuring the scanning mode), it is strongly recommended to set the polygon detection areas if you need to analyze only part of the frame.