Unexpected layout error Myriad VPU compute Stick 2

[VictorMrt]

Hello,

I have an issue when i'm using my model on Myriad neural compute stick 2. 

My model has 2 inputs : [1 128 1] and [1 8]

With the benchmark_app I can change the batch size by adding the following option :

-shape "inputs_cnn:0[2,128,1],inputs2:0[2,8]" (for size batch=2)

On CPU it works with no problem but when i launch it on the Myriad stick (-d MYRIAD) an error appears: 

[Step 1/11] Parsing and validating input arguments
[ INFO ] Parsing input parameters
[Step 2/11] Loading Inference Engine
[ INFO ] InferenceEngine:
API version ............ 2.1
Build .................. 2020.4.0-359-21e092122f4-releases/2020/4
Description ....... API
[ INFO ] Device info:
MYRIAD
myriadPlugin version ......... 2.1
Build ........... 2020.4.0-359-21e092122f4-releases/2020/4

[Step 3/11] Setting device configuration
[Step 4/11] Reading the Intermediate Representation network
[ INFO ] Loading network files
[ INFO ] Read network took 10.37 ms
[Step 5/11] Resizing network to match image sizes and given batch
[ INFO ] Reshaping network: 'inputs_cnn:0': [2, 128, 1], 'inputs2:0': [2, 8]
[ INFO ] Reshape network took 2.05 ms
[ INFO ] Network batch size: 2
[Step 6/11] Configuring input of the model
[Step 7/11] Loading the model to the device
[ ERROR ] Input inputs_cnn:0 has unexpected layout CHW

Do you know where it come from ?

Best regards, 

Victor 

[Syamimi_Intel]

Hi Victor Martin,

Thank you for reaching out. Based on your error, the main reason might be due to supported output layout.

Please have a try with using 4 dimensions (NCHW).

You can refer following link under supported output layout:

https://docs.openvinotoolkit.org/latest/openvino_docs_IE_DG_supported_plugins_Supported_Devices.html

In addition, you can also use the option of -b BATCH_SIZE to change the batch size:

https://docs.openvinotoolkit.org/2021.3/openvino_inference_engine_tools_benchmark_tool_README.html

Regards,

Syamimi

[Syamimi_Intel]

Hi Victor Martin,

This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.

Regards,

Syamimi