PUTvision / qgis-plugin-deepness

Deepness is a remote sensing plugin that enables deep learning inference in QGIS
https://www.buymeacoffee.com/deepness
Apache License 2.0
108 stars 25 forks source link

Error with Aerial Cars Detection model after upgrade to Deepness 0.6.4 #189

Closed hewag1975 closed 3 weeks ago

hewag1975 commented 1 month ago

Hi, thank you for the great plugin.

Describe the bug I recently updated to version 0.6.4 and since then the car detection using the Aerial Cars Detection model throws the following error:

image

To Reproduce Open QGIS and Deepness Plugin, add an image to analyze Select Input Layer and processed area Select Model Type "Detector" Load the model onxx-file

Expected behavior No error, model should be loaded to do inference

Screenshots (including Deepness options selected) image

Desktop (please complete the following information):

Thanks for your help.

bartoszptak commented 1 month ago

Hi @hewag1975 , could you perform these activities:

  1. Run QGIS
  2. Open Python Console (Ctrl+Alt+P)
  3. Type import onnxruntime, sys
  4. Type sys.version
  5. Type onnxruntime.__version__
  6. Type onnxruntime.get_build_info()
  7. Type onnxruntime.get_available_providers()
  8. Attach the screenshot with the versions.

It would be helpful for me to debug that issue.

PS: You can try to update the onnxruntime library version—the instruction you can find here. Just search the words "update the ONNXRuntime library in QGIS:".

hewag1975 commented 4 weeks ago

Hi @bartoszptak,

here is the output:

import onnxruntime, sys
sys.version
'3.9.18 (heads/master:5eba59e, Feb  1 2024, 20:02:10) [MSC v.1929 64 bit (AMD64)]'
onnxruntime.__version__
'1.17.0'
onnxruntime.get_build_info()
'ORT Build Info: git-branch=HEAD, git-commit-id=5f0b62cde5, build type=RelWithDebInfo, cmake cxx flags: /guard:cf /DWIN32 /D_WINDOWS /DWINAPI_FAMILY=100 /DWINVER=0x0A00 /D_WIN32_WINNT=0x0A00 /DNTDDI_VERSION=0x0A000000 /O2 /Ob1 /DNDEBUG /EHsc /EHsc /wd26812 -DEIGEN_HAS_C99_MATH -DCPUINFO_SUPPORTED'
onnxruntime.get_available_providers()
['AzureExecutionProvider', 'CPUExecutionProvider']

Updating the ONNXRuntime as you suggested resolves the problem

onnxruntime.__version__
'1.19.2'

Thanks. Feel free to close.