Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Adithya_N_Intel
Employee
105 Views

Unable to run smart classroom sample on VPU

Getting the below error while trying to run smart classroom demo with VPU on docker.

[ ERROR ] [ NOT_FOUND ] DVN_BATCH_ENABLED key is not supported for VPU

Openvino 2019 R1

Models used: face-detection-adas-0001, headpose-estimation-adas-0001, emotions-recognition-retail-0003, face-reidentification-retail-0095.

When googled with error, read somewhere that VPU has limited support for models. Could you confirm if that is the cause for this issue?

Please look into it and resolve.

TIA

0 Kudos
3 Replies
Luis_at_Intel
Moderator
105 Views

The models you have listed based on the documentation here have support for the Myriad plugins. 

  1. face-detection-adas-0001
  2. head-pose-estimation-adas-0001
  3. emotions-recognition-retail-0003
  4. face-reidentification-retail-0095

To make sure the models were working on the Myriad plugin, I tested the first 3 models using the Interactive Face Detection Demo (on a Windows 10 environment and a USB cam as input) and they seem to work as expected. 

Example:

<path-to>\inference_engine_samples_build\intel64\Release\interactive_face_detection_demo.exe -i cam -m "<path-to>\face-detection-adas-0001-fp16.xml" -m_ag "<path-to>\age-gender-recognition-retail-0013-fp16.xml" -m_hp "<path-to>\head-pose-estimation-adas-0001-fp16.xml" -m_em "<path-to>\emotions-recognition-retail-0003-fp16.xml" -d MYRIAD -d_ag MYRIAD -d_em MYRIAD -d_hp MYRIAD

 

After this I went ahead and tested the Smart Classroom Demo and for me it worked using the face-detection-adas-0001 and the face-reidentification-retail-0095 models as shown by the command below. May I ask how exactly is it that you are running the demo? If you can provide more details on your environment and steps followed to reproduce I can try to help resolve your issue. I hope you have found this information helpful, please let me know if you have any other questions. 

Example:

<path-to>\inference_engine_samples_build\intel64\Release\smart_classroom_demo.exe -m_act "<path-to>\person-detection-action-recognition-0005-fp32.xml" -m_fd "<path-to>\face-detection-adas-0001-fp16.xml" -m_reid "<path-to>\face-reidentification-retail-0095-fp16.xml" -fg "C:\Users\user\Documents\Intel\OpenVINO\openvino_models\faces_gallery.json" -i cam -d_act CPU -d_fd MYRIAD -d_reid MYRIAD

 

Regards,

@Luis_at_Intel

105 Views

Hi Luis

Thanks for your reply

We tried to run smart classroom demo in ubuntu 16.04.1 

We have faced the same issue as mentioned above 

Please find the attached png with the same.

Thanks

Aamna

Shubha_R_Intel
Employee
105 Views

Dear Syed, Aamna Iqra,

The main difference between what Luis did and what you tried is that your d_fd is HDDL. Can you try again with the latest OpenVino Release 2019R1.1 (released just this week) ? If you still suffer the same problem I will file a bug on your behalf.

Sorry for the inconvenience !

Shubha

Reply