debug code to testing in intel Devcloud but nothing happen

[christoven]

i follow this instruction to run video for testing my code on devcloud
https://jupyter.edge.devcloud.intel.com/user/u50944/notebooks/Reference-samples/iot-devcloud/openvino-2021.1/tutorials/object_detection/tutorial_object_detection_ssd.ipynb

i debug the code but nothing happen, no error message and not showing the result..

can help me to find the solution?
i try this tutorial to know the best specs for running mask and social distancing
thank you..

[Monique_J_Intel]

Hi @christoven,

Could you let me know what specific error you're seeing?

Kind Regards,

Monique Jones

[WaiMing_S_Intel]

Hi, I'm unable to access your link as I'm not an admin on DevCloud. Can you just check some basic debugging steps 

1. Were you able to successfully run any of the provided apps on a DevCloud system successfully with debugging messages shown? Without any changes to the default code
2. Were you able to successfully run a modified version of the provided apps on a DevCloud system successfully with debugging messages shown? 

Appreciate your input.

[christoven]

i got empty message when running the program. 
i never get any error message or error status. 

[WaiMing_S_Intel]

Have you tried some of the pre-installed scripts, are they working in your Jupyter environment?

[christoven]

i already can run the code and can build the job for testing specs in devloud with this code
#Submit job to the queue
job_id_core = !qsub main.sh -l nodes=1:idc001skl -F "results/core/ CPU FP32 {InputVideo} {NumRequests_CPU}" -N obj_det_core
print(job_id_core[0])
#Progress indicators
if job_id_core:
progressIndicator('results/core', f'pre_progress_{job_id_core[0]}.txt', "Preprocessing", 0, 100)
progressIndicator('results/core', f'i_progress_{job_id_core[0]}.txt', "Inference", 0, 100)
progressIndicator('results/core', f'post_progress_{job_id_core[0]}.txt', "Rendering", 0, 100)

but i got this new error message 

Unable to init server: Could not connect: Connection refused

(Detection Results:30144): Gtk-WARNING **: 01:00:11.716: cannot open display:

how can i fix this?
thank you ...

[Monique_J_Intel]

Hi Christoven, 

What are you expecting from the results? Are you looking to view it live? I suspect you are trying to write out to the console which isn't an easy task to do on the DevCloud. So if you just amend your code to write your results out to a video file and view them after the fact I think your issue will go away. 

Kind Regards,

Monique Jones

[christoven]

Hi..
my goal using intel devcloud is to know what is the best processor or specs for our mask detection + social distancing. 

on devcloud tutorial, for example: people_counter which i open and run people_counter.ipynb and click play button till the end of tutorial there is a part Submit to an edge compute node with an Intel® CPU

showing a process bar like this

after this process done, get files like below (this is tutorial, not our code)

output_64195.v-qsvr-2.devcloud-edge.mp4, 

post_progress_64195.v-qsvr-2.devcloud-edge.txt13 B6 minutes ago

stats_64195.v-qsvr-2.devcloud-edge.txt

now, on the next step when hit run button on part 4.7.0.1  View results from an Intel® CPU 

and choose the precessor i want or CPU i want to run, then i can see the result video with real people_counter.py 
i can choose what the best CPU or processor i want ,, lagging or not lagging, fps drop or normal fps thats our goal.

and my problem is when try with our code, the problem on this part  4.5.5.1  Submit to an edge compute node with an Intel® CPU

the message error i got was in the previous post.. 

thank you..

12 B6 minutes ago

4.45 MB5 minutes ago

[Monique_J_Intel]

Hi Christoven,

Thank you for the context and the code snippet. I see that you are doing imshow function call which will write the frame results to the console. You will need to comment this code out and replace it with code that will use OpenCV's video writer function like the code below that's referenced from the ONNX sample ->Reference-samples/iot-devcloud/openvino-2021.1/developer-samples/python/ONNX-object-detection-python/ONNX_object_detection.py :

#before inference loop

fourcc = cv2.VideoWriter_fourcc(*'avc1')
output_video = cv2.VideoWriter(output_path_file,fourcc, float(17.0), (640,360))

#At the end of the inference loop

output_video.write(postproc_frame)

Hope this helps. Also, planning to jump on a call to expedite your progress.

Kind Regards,

Monique Jones

[christoven]

Hi Monique
i want to try your suggestion but my access to devcloud is expired.

and i also need to inform my email register
my email registered under is aby@samelement.com

and please help my contents on devloud still same as before
thank you very much

[Monique_J_Intel]

Ok, I emailed you and we can go from there.

Kind Regards,

Monique Jones