Hey, I’m trying to use the Peoplenet model in my local setup, and I’ve successfully loaded the model using ONNX Runtime. The model returns two outputs with shapes (1, 12, 34, 60) and (1, 3, 34, 60). From my understanding, one represents the bounding box (bbox) and the other represents class confidence, with results corresponding to each grid cell. However, I’m not getting a clear explanation on how to post-process these outputs to extract the actual bounding boxes. Could you help with that?
Please provide complete information as applicable to your setup. Thanks
Hardware Platform (Jetson / GPU)
DeepStream Version
JetPack Version (valid for Jetson only)
TensorRT Version
NVIDIA GPU Driver Version (valid for GPU only)
Issue Type( questions, new requirements, bugs)
How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)
Could you also attach your whole pipeline you are using with DeepStream?
Sorry, I’m not running the PeopleNet V2 model in my DeepStream application. I’m only running the model on my local machine using ONNX Runtime in Python.
OK. Could you attach the link of the model you are using?
ngc registry model download-version “nvidia/tao/peoplenet:pruned_quantized_decrypted_v2.3.4”
Hi @sathishkumar.kannan , sorry for late reply. You can refer to our C/C++ source code deepstream\sources\libs\nvdsinfer_customparser\nvdsinfer_custombboxparser.cpp to learn how to postprocess the output layer of this model. Then implement the postprocessing in python yourself.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.