Deploying to Oak Camera pains

Hi!
I’ve got an issue where I’m trying to deploy my Roboflow model to my Oak 1 POE camera. I’ve tried the old way (Roboflow + Luxonis OAK), but that threw me an error.

File "/home/home-server/.local/lib/python3.10/site-packages/roboflowoak/__init__.py", line 63, in __init__
    int(self.model_objects["environment"]["RESOLUTION"]),
TypeError: int() argument must be a string, a bytes-like object or a real number, not 'list'

Then I tried the seemingly newer method here, which looks more promising: Deploy Models from Roboflow with the Luxonis DepthAI SDK.
When I try it on the standard models like the person-detection one, it runs fine, but when I run something else like mine (attached below), it returns this:

Downloading /home/home-server/.cache/blobconverter/cars-vqxmk_4_openvino_2021.4_6shave.blob...
{
    "exit_code": 1,
    "message": "Command failed with exit code 1, command: /opt/intel/openvino2021_4/deployment_tools/inference_engine/lib/intel64/myriad_compile -m /tmp/blobconverter/dd6d46240ed647558ef4da823c8a1971/cars-vqxmk_4/FP16/cars-vqxmk_4.xml -o /tmp/blobconverter/dd6d46240ed647558ef4da823c8a1971/cars-vqxmk_4/FP16/cars-vqxmk_4.blob -c /tmp/blobconverter/dd6d46240ed647558ef4da823c8a1971/myriad_compile_config.txt -ip U8",
    "stderr": "Unknown model format! Cannot find reader for model format: xml and read the model: /tmp/blobconverter/dd6d46240ed647558ef4da823c8a1971/cars-vqxmk_4/FP16/cars-vqxmk_4.xml. Please check that reader library exists in your PATH.\n",
    "stdout": "Inference Engine: \n\tIE version ......... 2021.4.2\n\tBuild ........... 2021.4.2-3974-e2a469a3450-releases/2021/4\n"
}
Closing OAK camera
Traceback (most recent call last):
  File "/home/home-server/Documents/Detection/oak_inference.py", line 16, in <module>
    oak.start(blocking=True)
  File "/home/home-server/.local/lib/python3.10/site-packages/depthai_sdk/oak_camera.py", line 286, in start
    self.build()  # Build the pipeline
  File "/home/home-server/.local/lib/python3.10/site-packages/depthai_sdk/oak_camera.py", line 380, in build
    component.on_init(self._pipeline, self._oak.device, self._pipeline.getOpenVINOVersion())
  File "/home/home-server/.local/lib/python3.10/site-packages/depthai_sdk/components/nn_component.py", line 129, in on_init
    self._blob = dai.OpenVINO.Blob(self._blob_from_config(self._config['model'], version))
  File "/home/home-server/.local/lib/python3.10/site-packages/depthai_sdk/components/nn_component.py", line 339, in _blob_from_config
    return blobconverter.from_openvino(xml=model['xml'],
  File "/home/home-server/.local/lib/python3.10/site-packages/blobconverter/__init__.py", line 478, in from_openvino
    return compile_blob(blob_name=Path(xml_name).stem, req_data={"name": Path(xml_name).stem}, req_files=files, **kwargs)
  File "/home/home-server/.local/lib/python3.10/site-packages/blobconverter/__init__.py", line 307, in compile_blob
    response.raise_for_status()
  File "/home/home-server/.local/lib/python3.10/site-packages/requests/models.py", line 953, in raise_for_status
    raise HTTPError(http_error_msg, response=self)
requests.exceptions.HTTPError: 400 Client Error: BAD REQUEST for url: https://blobconverter.luxonis.com/compile?version=2021.4&no_cache=False

Any idea what might be happening? Ideally a resolution to both would be good for all community users, but a solution for just one would still be fine.

Thank you in advance for your time!

  • **Project Type: Object Detection through Roboflow deployment on OAK 1 POE camera.
  • **Operating System & Browser: Ubuntu and Firefox
  • **Project Universe Link or Workspace/Project ID: https://app.roboflow.com/beast-detection/livestock-detection-vxbnv

Hi @The_Oracle - thanks for reaching out!

The OAK is a third-party integration and we cannot provide support for it. I recommend trying the Luxonis forum.

Sorry I couldn’t help more,
Jacob