Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.

Age-Gender model doesn't work


Hi there

I'm trying to use age-gender-recognition-retail-0013 model with OpenVINO 2020.1 on Movidius NCS v1 / Raspberry Pi 4 / Debian Buster.


When I run it with 2019 version of the model it fails with:

Traceback (most recent call last):
  File "./", line 44, in <module>
    _, confidences, boxes = net2.detect(frame, confThreshold=0.5)
cv2.error: OpenCV(4.2.0-openvino) ../opencv/modules/dnn/src/model.cpp:348: error: (-213:The function/feature is not implemented) Unknown output layer type: "SoftMax" in function 'detect'


When I run it with 2020 version of the model it fails with:

terminate called after throwing an instance of 'InferenceEngine::details::InferenceEngineException'
  what():  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_151](patternLabel_147: float{10,20,30}, patternLabel_148: int64_t{5}, patternLabel_150: int64_t{1}) -> (??)':
The axis must => 0 and <= input_rank (axis: 4294967295).

Attached are the two python files - one for 2019 and one for 2020 model file.

Any idea how to fix it?

Thanks! Michael

0 Kudos
1 Reply

Hi Michael,

Thanks for reaching out.

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 issue you can do either of these options:

  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.
  3. Install the previous version of OpenVINO™ toolkit (2019 R3.1) on a Windows, Linux or Mac OS system, convert the model there, which will generate the IR v7 format and that should work successfully with the Intel® Neural Compute Stick 2 on the Raspberry Pi.

Best regards,


0 Kudos