Your problem seems to originate from directories access.
I believe you are running this code with Anaconda.
As a workaround, try to run the Anaconda cmd as admin & run jupyter notebook in it.
Then, try to see whether this conflict still persist.
Btw, you need to fill in the --saved_model_dir <location of your saved model> and --input_shape <write your input shape>
I did try running the command from the terminal as admin as well still getting the same error, decided to do this on the intel dev cloud, this time i was able open the directory but this is what i got
My tensor flow saved model directory name is 'model' hence I pass model to --saved_model_dir
I used the command that can be found in OpenVino Official Documentation:
This is what i get:
Convert yolo to IR
python mo_tf.py --input_model C:\Users \Desktop\OpenVINO-Experiments\Yolov3\tensorflow-yolo-v3\frozen_darknet_yolov3_model.pb --tensorflow_use_custom_operations_config C:\Users \Desktop\OpenVINO-Experiments\Yolov3\yolo_v3.json --batch 1
And I'm able to Run object_detection_demo_yolov3_async.py
python object_detection_demo_yolov3_async.py -i inputVideo.mp4 -m frozen_darknet_yolov3_model.xml -d CPU
Perhaps you can do the same by referring to the official documentation
This is a video that might help you: https://www.youtube.com/watch?v=FaqVhvJ6-Uc
I am using tensorflow 2.2, therefore my model is in SavedModel format which consists of a directory with a saved_model.pb file and two subfolders: variables and assets.
According to the documentation i am requred to give this command
python3 mo_tf.py --saved_model_dir <SAVED_MODEL_DIRECTORY>
I have successfully run the command but im getting this error
Okay found a solution, rather the reason that was causing the problem.
So the thing is I was using openVINO 2020.3 toolkit this whole time.
I have upgraded to openVINO 2020.4 and now all the problems have gone.
If anyone is facing this issue I hope this helps.