Since you're using Tensorflow, you'll have to change your code to declare a specific size for the input placeholder instead of using a variable size (variable sized inputs are set as "none"). Here is an example of a fixed input size:
The software kit that is compatible with the NCS2 is OpenVINO. Here is the installation page for OpenVINO on Raspbian. Please let me know if you have any more questions!
I was able to reproduce this issue. Can you provide the Keras model and the steps you made to convert the model to Tensorflow and then to NCS graph?
Thanks so much for your patience!
This might even be a USB issue - are you using a USB hub or are you plugging directly into your system? (the xlink error is usually due to communication issues between the NCS & the system) since it does look like you converted …
Your graph file is likely corrupt since you're using a previous version of the NCSDK. Reinstall the latest version of the NCSDK here. And then try to recompile the graph file.
Let me know if this was helpful
Yes! You can use multiple NCS devices to speed up inference. Are you using the NCSDK with the NCS1 or OpenVINO?
If you're using the NCSDK, there's a GoogLeNet sample in the ncappzoo that shows how to use more than one NCS (here). Ano…
Thanks for reaching out! Some questions for you:
1. Are you using the NCS1 or NCS2? I'm not able to tell by the error messages, but sometimes you can get that error if your're using an NCS2 with the NCSDK (NCSDK is not compatibl…
If you have already built your image classifier, you can try following these instructions to run inference on the NCS. Please let me know if you have any further questions!
It looks like you're trying to use the NCSDK with the NCS2. The NCSDK is only compatible with the NCS1, not the NCS2. The software kit that is compatible with the NCS2 is OpenVINO and can be installed here . I hope this informat…
I apologize if I misunderstood your question, but are you asking how to open the NCS device in your own code? If so, here is a code snippet (in C++ on how you would open the NCS device) so you can add this in your code to call the NCS…
I can help you out, but first can you provide more details about your device/system you're using please? Are you using the NCS1 or NCS2? What system are you running on?
Try enabling the debug option in the Tensorflow parser - this could provide more information about the error and which layer/operation is throwing it. You will need to set the debug flag to True in the following file as sudo:
I just wanted to know how the performance of your NCS2 compares to others to make sure there's no other issues. You can try the security_barrier_camera_demo, interactive_face_detection_demo, segmentation_demo and I can compare your per…
Did you create filters for your virtual machine when installing? Here is the link for instructions on how to do that (this page is for installing the ncsdk on a virtual machine, but the process is almost the same for OpenVINO…
I think this is caused by mixed versions of Numpy installed. Have you tried uninstalling then reinstalling Numpy? Are you getting a "cannot import name 'validate_lengths' error with v1.16? This Github thread might be helpful to you.…
The NCS is only able to run CNNs. In theory, however, we should be able to support MLPs as long as you're using an image input. I have not tried to compile an MLP myself, so I cannot confirm this. But we do support networks with MLP in …
I apologize for any confusion. But you're correct - the NCSDK is not compatible with the NCS2. The software kit that is used with the NCS2 is called OpenVINO and can be installed here . Here is a good external blog about how to install …
Since you're using the NCS2, you will need to use the OpenVINO toolkit. You should be able to convert your Tensorflow FaceNet model using the Model Optimizer included in the toolkit. Take a look at the following section in the documen…