Daniil-Osokin / lightweight-human-pose-estimation-3d-demo.pytorch

Real-time 3D multi-person pose estimation demo in PyTorch. OpenVINO backend can be used for fast inference on CPU.
Apache License 2.0
653 stars 137 forks source link

Where is human-pose-estimation-3d.xml #58

Closed riboyuan99 closed 3 years ago

riboyuan99 commented 3 years ago

After I ran "python /deployment_tools/model_optimizer/mo.py --input_model human-pose-estimation-3d.onnx --input=data --mean_values=data[128.0,128.0,128.0] --scale_values=data[255.0,255.0,255.0] --output=features,heatmaps,pafs"

then ran "python demo.py --model human-pose-estimation-3d.xml --device CPU --use-openvino --video 0",

it gives me an error said "Exception: Path to the model human-pose-estimation-3d.xml doesn't exist or it's a directory"

What did I do wrong?

Daniil-Osokin commented 3 years ago

Hi! Looks like model was not converted by model optimizer. What is the output of the first command (have you substituted with a real path to OpenVINO install directory)?

riboyuan99 commented 3 years ago

Thanks for the reply! The output is: zsh: no matches found: --mean_values=data[128.0,128.0,128.0]

Screen Shot 2021-03-02 at 3 43 39 PM
Daniil-Osokin commented 3 years ago

You see, there is the error: no matches found: --mean_values=data[128.0,128.0,128.0], so model optimizer did not finish properly. Looks like you need to escape this block with quotes "data[128.0,128.0,128.0]", check this issue.

riboyuan99 commented 3 years ago

Thank you very much! I will try that later.