[ ERROR ] Check 'axis < static_cast<size_t>(input_rank)' failed

[Xiong__Zhuang]

When I run the OpenVINO_2020_1 sample by Raspberry Pi with NCS 2:

./armv7l/Release/object_detection_sample_ssd -m face-detection-adas-0001.xml -d MYRIAD -i /home/pi/Pictures/1.jpg

then

 Check 'axis < static_cast<size_t>(input_rank)' failed at /teamcity/work/scoring_engine_build/releases_2020_1/ngraph/src/ngraph/op/gather.cpp:140:
While validating node 'Gather[Gather_852](patternLabel_848: float{10,20,30}, patternLabel_849: int64_t{5}, patternLabel_851: int64_t{1}) -> (??)':
The axis must => 0 and <= input_rank (axis: 4294967295).

How to solve it?

[David_C_Intel · Accepted solution]

Hi Zhuang,

Thanks for reaching out.

Could you please test the sample with the previous version of the model and let us know if it works?

https://download.01.org/opencv/2019/open_model_zoo/R3/20190905_163000_models_bin/face-detection-retail-0004/FP16/face-detection-retail-0004.xml

https://download.01.org/opencv/2019/open_model_zoo/R3/20190905_163000_models_bin/face-detection-retail-0004/FP16/face-detection-retail-0004.bin

Best Regards,

David

[Xiong__Zhuang]

When I test the sample you provide, it works. What's the reason?

DavidC (Intel) wrote:

Hi Zhuang,

Thanks for reaching out.

Could you please test the sample with the previous version of the model and let us know if it works?

https://download.01.org/opencv/2019/open_model_zoo/R3/20190905_163000_mo...

https://download.01.org/opencv/2019/open_model_zoo/R3/20190905_163000_mo...

Best Regards,

David

[David_C_Intel]

Hello Zhuang,

Thank you for your reply.

This is an issue with the latest models due to changes in the Model Optimizer and ngraph. The development team is looking into it, we will let you know once we have an update.

As a workaround for this issue, you can use the models from the OpenVINO™ toolkit 2019 R3.1 release.

Best Regards,

David

[David_C_Intel]

Hi Zhuang,

This is an issue with the latest OpenVINO™ toolkit version (2020.1). The "axis" error is caused by the new IR v10 format. 

As a workaround for this you can:

1. Use the previous pre-trained models from the 2019 R3 open model zoo.
2. Generate the IRv7 format files using the Model Optimizer from the full installation of OpenVINO™ toolkit with the following parameter: --generate_deprecated_IR_V7

Best Regards,

David

[teng__hsiang_yuan]

Hi I met similar problem, too.

I use another open model , but it can't work. 

How can I fix it?

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>------------------>>>>>>>>>>>>>>>>>>>>>>>

EX:

./armv7l/Release/object_detection_sample_ssd -m person-vehicle-bike-detection-crossroad-0078.xml -d MYRIAD -i ~/Pictures/1.jpg

Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ]     /home/pi/Pictures/1.jpg
[ INFO ] Loading Inference Engine
[ INFO ] Device info:
        MYRIAD
        myriadPlugin version ......... 2.1
        Build ........... 37988
[ INFO ] Loading network files:
        person-vehicle-bike-detection-crossroad-0078.xml
        person-vehicle-bike-detection-crossroad-0078.bin
[ ERROR ] Check 'axis < static_cast<size_t>(input_rank)' failed at /teamcity/wor            k/scoring_engine_build/releases_2020_1/ngraph/src/ngraph/op/gather.cpp:140:
While validating node 'Gather[Gather_7173](patternLabel_7169: float{10,20,30}, p            atternLabel_7170: int64_t{5}, patternLabel_7172: int64_t{1}) -> (??)':
The axis must => 0 and <= input_rank (axis: 4294967295).

[Cohen__Mickey]

I had the same problem. Downloaded the face models and now I have:

python3.7 run.py -s 0 --ir ./face-detection-retail-0004.xml

Starting application...
   - Camera Source: 0
   - Plugin:       Myriad
   - IR File:       ./face-detection-retail-0004.xml
   - Input Shape:   [1, 3, 300, 300]
   - Output Shape:  [1, 1, 200, 7]
   - Labels File:   ../../data/ilsvrc12/synset_labels.txt
   - Mean File:     None

 Press any key to exit.

Traceback (most recent call last):
  File "run.py", line 204, in <module>
    sys.exit(main())
  File "run.py", line 180, in main
    res_label, res_conf, e_time = perform_inference(input_img, exec_net, input_blob, output_blob)
  File "run.py", line 128, in perform_inference
    predicted_label = label_list[int(top_ind[0])]
TypeError: only size-1 arrays can be converted to Python scalars

[PJIA01]

• Same problem here. But, yes, --generate_deprecated_IR_V7 works for me.... Thank you DavidC.
• And, for me, I've got an extra ERROR on Raspberry Pi 4.
  • On my laptop, object_detection_sample_ssd runs fine.
  • On Raspberry Pi 4, object_detection_sample_ssd_c runs fine.
  • On Raspberry Pi 4, object_detection_sample_ssd failed.

➜ samples ./cpp/build/armv7l/Release/object_detection_sample_ssd -m face-detection-adas-0001.xml -d MYRIAD -i me.jpg
[ INFO ] InferenceEngine:
API version ............ 2.1
Build .................. custom_releases/2020/1_d349c3ba4a2508be72f413fa4dee92cc0e4bc0e1
Description ....... API
Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ] me.jpg
[ INFO ] Loading Inference Engine
[ INFO ] Device info:
MYRIAD
myriadPlugin version ......... 2.1
Build ........... 37988
[ INFO ] Loading network files:
face-detection-adas-0001.xml
face-detection-adas-0001.bin
[ INFO ] Preparing input blobs
[ INFO ] Batch size is 1
[ INFO ] Preparing output blobs
[ INFO ] Loading model to the device
[ INFO ] Create infer request
[ WARNING ] Image is resized from (924, 1280) to (672, 384)
[ INFO ] Batch size is 1
[1] 9179 segmentation fault ./cpp/build/armv7l/Release/object_detection_sample_ssd -m -d MYRIAD -i me.jp

Raspberry Pi 4 hangs and quit right away...

Anybody please investigate this issue a bit?

Thank you very much

Pei

[David_C_Intel]

Hi Pei,

I noticed you also opened another similar thread here, do you mind if we continue troubleshooting on that thread?

Best regards,

David

[Tr__Ra]

Hello,

I am still getting the same error the OP mentioned a few months ago.

Now that 20.02 is released, any progress on this error?

Thank you

[David_C_Intel]

DavidC (Intel) wrote:

Hi Zhuang,

This is an issue with the latest OpenVINO™ toolkit version (2020.1). The "axis" error is caused by the new IR v10 format. 

As a workaround for this you can:

1. Use the previous pre-trained models from the 2019 R3 open model zoo.
2. Generate the IRv7 format files using the Model Optimizer from the full installation of OpenVINO™ toolkit with the following parameter: --generate_deprecated_IR_V7

Best Regards,

David

Hello Ra,

The latest OpenVINO™ toolkit version 2020.2  for Raspbian is yet to be released. The "axis" error will be fixed on that version, for now, as a workaround you can follow the steps described before.

Regards,

David

[Pedreira_Gonzalez__I]

How can we do option 2? If the model is in OpenvinoFormat? Can you share a model_optimizer sample command line to do that¿

Thank you

DavidC (Intel) wrote:

Quote:

DavidC (Intel) wrote:

Hi Zhuang,

This is an issue with the latest OpenVINO™ toolkit version (2020.1). The "axis" error is caused by the new IR v10 format. 

As a workaround for this you can:

1. Use the previous pre-trained models from the 2019 R3 open model zoo.
2. Generate the IRv7 format files using the Model Optimizer from the full installation of OpenVINO™ toolkit with the following parameter: --generate_deprecated_IR_V7

Best Regards,

David

Hello Ra,

The latest OpenVINO™ toolkit version 2020.2  for Raspbian is yet to be released. The "axis" error will be fixed on that version, for now, as a workaround you can follow the steps described before.

Regards,

David