Inference time better when using NCS1 in comparison with NCS2

[UPetr]

I am running OpenVino R2 2019 on Raspberry Pi 3 B+ and performed inference using pre-trained intel model 'person-detection-retail-0013'.

The recorded approximate inference time using NCS1 was 1.94 fps, while using  NCS2 1.93 fps.

Code snippets:

...
# Specify target device
self.net.setPreferableTarget(cv.dnn.DNN_TARGET_MYRIAD)
...

# Prepare input blob and perform an inference.
blob = cv.dnn.blobFromImage(frame, size=(672, 384), ddepth=cv.CV_8U)
self.net.setInput(blob)
out = self.net.forward()

What is the problem here, NCS2 should be much faster? Is it Raspberry pi USB bottleneck, usage of OpenVino's R2 version or something else?

Regards,

Uros

[JesusE_Intel]

Hi Uros,

Thanks for reaching out. It is possible that the USB ports of Raspberry Pi, don't have the power required for the NCS2, I recommend you use a power USB hub. Also, updating to the latest version (2020.1.023) of OpenVINO™ Toolkit for Raspbian* OS and using the Inference Engine instead of the OpenCV DNN library.

Regards,

Jesus

[UPetr]

Hi Jesus,

Thanks for your answer.

According to the official RPI specifications: The Raspberry Pi 3 B+ can supply 1.2 A to downstream USB peripherals. 

Since I am not using any other USB peripherals except NCS2 how is that not enough to run it properly? Can you give me some power/current requirements for NCS2?

Regards,

Uros

[JesusE_Intel]

Hi Uros,

You should be able to find the specs on the Platform Configurations documentation.

I did a quick test with the Intel Movidius Neural Compute Stick (Intel Movidius NCS) and the Intel NCS 2 with the benchmark_app and person-detection-retail-0013 model.

With the model mentioned above, I got the following results using the OpenVINO toolkit 2020.1 for Raspbian OS pre-built package.. Please note that results will very depending on the model.

Intel Movidius NCS: 7.04 FPS

Intel NCS 2: 13.70 FPS.

I used the benchmark_app and the following command.

./benchmark_app -d MYRIAD -m person-detection-retail-0013.

Hope this helps!

Regards,

Jesus