Community
cancel
Showing results for
Did you mean:
Beginner
371 Views

## Do you have sample models (xml/bin files) for the sample exes (classification_sample_async.exe) provided ? Also missing dll errors during execution of the exe

Hi,

I am using the NCS-2 for teaching and developing some applications. Both the sample demos (demo_squeezenet_download_convert_run.bat and demo_security_barrier_camera.bat) run fine.

1. Attempting to run the sample binary classification_sample_async.exe from C:\Users\<MyUserName>\Documents\Intel\OpenVINO\inference_engine_samples_build\intel64\Debug causes a few missing dll errors, namely, inference_engined.dll in C:\Program Files (x86)\IntelSWTools\openvino\inference_engine\bin\intel64\Debug, tbb_debug.dll in C:\Program Files (x86)\IntelSWTools\openvino\inference_engine\bin\intel64\Debug and opencv_imgcodecs412d.dll in C:\Program Files (x86)\IntelSWTools\openvino\opencv\bin whereas I find that these dlls are within the Intel OpenVINO installation itself. I added these to the system path explicitly and then the exe didn't complain. I would need to manually change these paths depending on Debug/Release though. Is this correct or is there an easier automated way ?
2. However, for inferencing, I need to provide it with sample models (.xml/bin files), I don't see any sample .xml files relevant to the exe. Are there any sample xml and bin files that I can use off-the-shelf to run the exe ?
3. Building individual sample applications on Windows causes a CMake error namely, "Unknown CMake command "ie_add_sample" but building all the applications succeeds ! I googled and came across the Intel link https://forums.intel.com/s/question/0D50P00004RxnLg/cmake-errors-of-interactivefacedetectiondemo?lan... -- seems its a known issue as early as 4 months back. Is there a permanent fix yet ?

My environment details:-

Windows 10 Pro running on Intel i7-8550U (64 bit)

Intel OpenVINO_2019_3.379

Python ver 3.7.2

Cmake ver 3.16.2

Moderator
104 Views

Thank you for reaching out.

1- You will need to change these paths manually, as the system does not know if you want to add them in your command or not.

There you can find the binary file and run the following commands:

`python3 downloader.py --help`

• Install the pre-trained model called "MODEL_NAME" in the OUTPUT_DIRECTORY specified:
`python3 downloader.py --name MODEL_NAME -o OUTPUT_DIRECTORY`

3- To build the demos you will need to use build_demos.sh script located in openvino>/deployment_tools/open_model_zoo/demos directory. The script will build all the demos and print the location of the binaries.

Regards,

Mauricio R.