Can't run Classification Python Demo with OpenVINO 2022.1

[Mostafa-]

Hello All,

I installed openvino-dev[caffe] successfully, and download squeesenet1.1 by;

omz_downloader --name sqeezenet1.1

 then convert it successfully by;

mo --input_model public/squeezenet1.1/squeezenet1.1.caffemodel --batch 1 --output_dir squeezenet_IR

then clone open_model_zoo to use the classification python demo;

python3 classification_demo.py -m /home/mostafa/intel/squeezenet_IR/squeezenet1.1.xml -i classification.gif --labels /home/mostafa/intel/open_model_zoo/data/dataset_classes/imagenet_2015.txt 

and I got the below error;

Traceback (most recent call last):
  File "classification_demo.py", line 28, in <module>
    from openvino.model_zoo.model_api.models import Classification, OutputTransform
ModuleNotFoundError: No module named 'openvino.model_zoo.model_api'

 Thanks,

Mostafa

[Mostafa-]

I installed also openvino-dev[tensorflow2] and tried also object detection demo - python - with tiny yolov4 and I got the same error;

Traceback (most recent call last):
  File "object_detection_demo.py", line 28, in <module>
    from openvino.model_zoo.model_api.models import DetectionModel, DetectionWithLandmarks, RESIZE_TYPES, OutputTransform
ModuleNotFoundError: No module named 'openvino.model_zoo.model_api'

[Vladimir_Dudnik]

@Mostafa- starting from 2022.1 OMZ most of python demos require to install an additional module, for OMZ Model API, see documentation for more details

Installing Python* Model API package

Use the following command to install Model API from source:

pip install <omz_dir>/demos/common/python

[Mostafa-]

Thank You Vladimir.

I did it and there's another error;

(openvino-tensorflow_env) mostafa@mostafa-VirtualBox:~/intel/open_model_zoo/demos/object_detection_demo/python$ python3 object_detection_demo.py --model tiny-yolov4/FP32/yolo-v4-tiny-tf.xml -at yolo -i test.mp4 -t 0.5 -d CPU -o fp32_output.avi
[ INFO ] OpenVINO Runtime
[ INFO ] build: 2022.1.0-7019-cdb9bec7210-releases/2022/1
[ INFO ] Reading model tiny-yolov4/FP32/yolo-v4-tiny-tf.xml
[ WARNING ] The parameter "input_size" not found in YOLO wrapper, will be omitted
[ WARNING ] The parameter "num_classes" not found in YOLO wrapper, will be omitted
[ INFO ] Input layer: image_input, shape: [1, 416, 416, 3], precision: f32, layout: NHWC
[ INFO ] Output layer: conv2d_17/BiasAdd:0, shape: [1, 13, 13, 255], precision: f32, layout: 
[ INFO ] Output layer: conv2d_20/BiasAdd:0, shape: [1, 26, 26, 255], precision: f32, layout: 
[ INFO ] The model tiny-yolov4/FP32/yolo-v4-tiny-tf.xml is loaded to CPU
[ INFO ] Device: CPU
[ INFO ] Number of streams: 1
[ INFO ] Number of threads: AUTO
[ INFO ] Number of model infer requests: 2
Traceback (most recent call last):
  File "object_detection_demo.py", line 304, in <module>
    sys.exit(main() or 0)
  File "object_detection_demo.py", line 205, in main
    results = detector_pipeline.get_result(next_frame_id_to_show)
  File "/home/mostafa/intel/openvino-tensorflow_env/lib/python3.8/site-packages/openvino/model_zoo/model_api/pipelines/async_pipeline.py", line 124, in get_result
    result = self.model.postprocess(raw_result, preprocess_meta), {**meta, **preprocess_meta}
  File "/home/mostafa/intel/openvino-tensorflow_env/lib/python3.8/site-packages/openvino/model_zoo/model_api/models/yolo.py", line 122, in postprocess
    detections = self._parse_outputs(outputs, meta)
  File "/home/mostafa/intel/openvino-tensorflow_env/lib/python3.8/site-packages/openvino/model_zoo/model_api/models/yolo.py", line 224, in _parse_outputs
    detections += self._parse_yolo_region(out_blob, meta['resized_shape'], layer_params[1])
  File "/home/mostafa/intel/openvino-tensorflow_env/lib/python3.8/site-packages/openvino/model_zoo/model_api/models/yolo.py", line 130, in _parse_yolo_region
    predictions = permute_to_N_HWA_K(predictions, params.bbox_size, params.output_layout)
  File "/home/mostafa/intel/openvino-tensorflow_env/lib/python3.8/site-packages/openvino/model_zoo/model_api/models/yolo.py", line 44, in permute_to_N_HWA_K
    tensor = tensor.reshape(N, -1, K, H, W)
ValueError: cannot reshape array of size 43095 into shape (1,newaxis,85,13,255)

[IntelSupport · Accepted solution]

Hi Mostafa,

The error is because you wrongly specify the architecture_type of yolo-v4-tiny-tf model. The architecture_type of the model should be yolov4. You can refer to the Supported Models for the architecture_type of each model. I have verified the demo and the output is as below:

Command used :

object_detection_demo.py -i <path_to_video>\car-detection.mp4" -m "<path_to_model>\yolo-v4-tiny-tf.xml" -at yolo v4 -t 0.5 -o object_detection_output.avi

Regards,

Aznie

[Mostafa-]

thank you, it works now

[IntelSupport]

Hi Mostafa,

This thread will no longer be monitored since this issue has been resolved. If you need any additional information from Intel, please submit a new question. 

Regards,

Aznie