The VA Face Recognition Pack add-on (see Installing DetectorPack add-ons) includes a utility that allows you to create GPU neural network caches without using Axxon One. The presence of the cache speeds up the initialization, optimizes video memory consumption, and ensures the start and correct operation of the Face detector VA on the GPU.

To optimize the operation of the Face detector VA on GPU, do the following:

1. Stop the server (see Stop the server).
   

   If the system uses the software that is running on the GPU, you must stop its operation.

2. Create the GPU_CACHE_DIR system variable (see Creating system variable). In the Variable value field, specify the path to the cache location in Latin characters only. For example, D:\GPU_cache. The specified directory will store the cache for all used detectors and neural networks.
   

3. Run the command line as administrator.

4. Starting with Detector Pack 3.14:

   1. Call the utility, specifying the required video stream resolutions in the XxX or XxX XxX format via the -r key. In this case you must not further specify the video stream resolution.
      Example:
      

      C:\Program Files\Common Files\AxxonSoft\DetectorPack\VaFaceGpuCacheGenerator.exe -r 3840x2160 2688x1520

      If you don't specify the resolution explicitly, the utility uses the default 1920x1080 resolution for creating the cache.
      

      To get help information, you can enter the query in the command line: C:\Program Files\Common Files\AxxonSoft\DetectorPack\VaFaceGpuCacheGenerator.exe --help

   2. Press Enter.

   3. Specify the ID of the required GPU (see Selecting Nvidia GPU when configuring detectors).

   4. Press Enter. As a result, the process of creating the cache starts. If you get the "0" value, it means that the cache is created successfully.
      

5. For Detector Pack 3.13 and older:
   1. In the command line, enter the C:\Program Files\Common Files\AxxonSoft\DetectorPack\VaFaceGpuCacheGenerator.exe query to call the utility.

   2. Press Enter.

   3. Specify the ID of the required GPU (see Selecting Nvidia GPU when configuring detectors).

   4. Press Enter.

   5. If you want to change the resolution of the video stream:

      1. Enter y. If you don't want to change the resolution of the video stream, enter n. 

      2. Specify the required resolution. For example, 3840x2160 2688x1520. The GPU uses the default 1920x1080 video stream resolution. If you change the resolution of the video stream from 1920x1080 and higher, it increases the percentage of face detection.

6. Press Enter. As a result, the process of creating the cache starts. If you get the 0 value, it means that the cache is created successfully.
   

Optimizing the operation of the Face detector VA on the GPU is complete.