I have several openvino models in a program.
On running anyone model the code runs fine but when I try to run another model it gives me this error. Which makes me belive there's a certain way to load/unload a network or model which should not give this error.
Since the documentation says you can run load any number of netwroks this should not be a problem.
Can you help here?
Have you tried running any of the demos. If not, Can you try with interactive_face_detection_demo which uses more than two models and let us know the behavior. How many models are you using and what are they?
I tried the demo and it works fine. What i am trying to do is have an app where someone can select which ML model they want to run and then run that.
So on runnning the app, whenever I choose an openvino model and run, it works fine but when I close that model and run another model, it keeps giving me this error.
I am running pose estimation, object detection and face detcetion. But none of them together, one after the other.
The issue seems to occur as the MYRIAD is still being used by the previous model. Have you referred to any OpenVINO demos to load the models?
If you are still facing the issue, could you please provide the necessary files so that we can have a look. If required, I can send you a PM to share the files privately.
I have faced similar issues. Please give some time to NCS to relax as well as loading and Unloading takes time depending on model and its optimization. Use time.sleep(5) or use try catch in loop.
I Hope, It will help you