Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU): Jetson AGX Orin
• DeepStream Version 7.1
• JetPack Version (valid for Jetson only) 6.1
• TensorRT Version 10.3.0.30
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs) Question
• 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)
My goal is to use a custom pre-trained model to extract features for each detected person from an input stream with DeepStream. I currently have little to none idea how to do it and would like to have some guidances. Here are my current situation:
- I have the onnx version of the model (converted from .pth) => Do I need to implement a script to convert it to a tensorrt engine or the conversion will be automatically done in the device? If an implemented conversion is required, how should I do it (I am currently looking at this repo but still have no idea how to do it for my custom model)?
- Next, I followed the README inside the deepstream folder located at /opt/nvidia/deepstream/deepstream-7.1/sources/tracker_ReID/README to download the TAO pre-trained model and run it with NvDeepSort (instead of NvDCF_perf) in deepstream_test2. This README also have a brief instruction of how to run a custom model for Re-ID. I haven’t tried that but it seems like all I need to modify the tracker config file?
Thanks a lots!