OpenPose Trt engine convertion failed

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 6.0
• JetPack Version (valid for Jetson only)
• TensorRT Version 8.1
• NVIDIA GPU Driver Version (valid for GPU only) 470

I tried to convert openpose caffee model in deepstream, i mentioned output-blob-names=‘concat_stage7’ but when converting to trt engine i am getting error says there is no such layer.
i have checked openpose repo there the same layer they are using as output layer.

ERROR:
ERROR: …/nvdsinfer/nvdsinfer_model_builder.cpp:163 Could not find output layer '‘concat_stage7’

3

could you share a repo or more info? or can you search the openpose DeepStream implementation in Ethernet?

4

Most of the implementations are in c/c++, but i want to run using python3. Also i did not found a better documentation for writing custom parser function and compile.

5

Confusing…
do you want us to help the model parser error you orginally asked, or you have fixed it and now have another question?

6

need help to write parser function

7

Could you refer to the post_processor under GitHub - NVIDIA-AI-IOT/deepstream_tao_apps: Sample apps to demonstrate how to deploy models trained with TAO on DeepStream ?

We don’t know what the meaning of the network output, so you could refer to above sample to add a network parser callback to parse the network output.

9

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.