Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Beginner
57 Views

Slow FPS on Neural Compute Stick 2 (Myriad X) with Mobilenet SSD

Hi There, 

We have tried the official object_detection_sample_ssd (shipped along with openVINO R4), and compared the perf between NCS (Myriad2) and NCS2 (Myriad X), we found the perf gain is very limited. Say for NCS (Myriad2) is about 5.8 fps, and for NCS2 (MyriadX) is about 6.3 fps. Does anyone know if any settings can be tuned? Since NCS2 is said to have about at least 3x perf on this network.

Any hints will be highly appreciated~

0 Kudos
7 Replies
Highlighted
Beginner
57 Views

Hello, anyone help to answer this question please? Since I did see intel show ssd-mobilenet demo with ncs2, which was able to run at around 30fps, on the intel AIDC event.

0 Kudos
Highlighted
Beginner
57 Views

Just trying to popup this thread, see if anyone could help on this. thanks

0 Kudos
Highlighted
57 Views

Where does your ssd-mobilenet model comes from?  I tried the mobilenet-ssd downloaded by using model_downloader. That network is trained by using caffe and its input resolution is 300x300.  Its performance on ncs2 is expected on my env,  about 30 fps when running with 'benchmark_app' in the package. 

0 Kudos
Highlighted
57 Views

same here. I am going to try it with the new openvino R5 ...

0 Kudos
Highlighted
Beginner
57 Views

The model of your own training, the computing power is similar to mobilenet_ssd, but the input is 512*512. Is NCS2 particularly sensitive to the size of the model input?

0 Kudos
Highlighted
Employee
57 Views

Dear folks,

Please upgrade to 2019 R1 if you haven't already.

Thanks for using OpenVino !

Shubha

0 Kudos
Highlighted
Beginner
57 Views

Hi, I have a problem. Using SSD Mobilenet v1 with 90 classes_num in config I get 0.05 seconds inference time, however , when I use my own trained model with only 1 classes_num I get 4.7 seconds inference time. Have I missed something. I use ssd_v2_support.json for model optimization.

0 Kudos