MYRIAD device, error: Dims and format are inconsistent

hake__malte · ‎08-07-2019

Hi, for batch size > 1i get the error "Failed to call QueryNetwork for MYRIAD device, error: Dims and format are inconsistent."

The optimized model is the Tensorflow slim model resnet_v2_50.

[ INFO ] InferenceEngine:

API version ............ 2.0

Build .................. custom_releases/2019/R2_f5827d4773ebbe727c9acac5f007f7d94dd4be4e

Description ....... API

[ INFO ] Parsing input parameters

[ INFO ] Files were added: 2

[ INFO ] /home/malte/val2017_test/000000007574.jpg

[ INFO ] /home/malte/val2017_test/00000000757.jpg

[ INFO ] Creating Inference Engine

GPU

clDNNPlugin version ......... 2.0

Build ........... 27579

HETERO

heteroPlugin version ......... 2.0

Build ........... heteroPlugin

MYRIAD

myriadPlugin version ......... 2.0

Build ........... 27579

[ INFO ] Loading network files

[ INFO ] Preparing input blobs

[ WARNING ] Image is resized from (640, 480) to (224, 224)

[ INFO ] Batch size is 2

[ INFO ] Loading model to the device

[ ERROR ] Failed to call QueryNetwork for MYRIAD device, error: Dims and format are inconsistent.

Can you tell me what is wrong here?

Thanks,

Malte

Shubha_R_Intel · ‎08-08-2019

Dear hake, malte,

Please give me more information about your model. Also please give me the model optimizer command you used as well. I see that you are using OpenVino 2019 R2 (from API version ............ 2.0) so that is good.

What happens when you run your code on CPU ?

Thanks,

Shubha