Segmentation fault (core dumped)

[yhwang-hub]

When I run the command ./deepstream -s file:///home/wyh/DeepStream-Yolo-Pose/bodypose.mp4 -c config_infer_primary_yoloV8_pose.txt -w 1280 -e 720 in the docker of deepstream6.2, an error occurs.
Segmentation fault (core dumped)

[marcoslucianops]

Please use the correct exporter file (from utils) to export the model.

[teddybee]

I have similar problem. The root cause could be I cant export the onnx model from yolo11l-pose.pt. Following the documentation I got this error:

## Exception summary

<class 'AttributeError'>: 'float' object has no attribute 'node'

While executing %item : [num_users=1] = call_function[target=torch.ops.aten.item.default](args = (%getitem_24,), kwargs = {})
Original traceback:
File "/root/ultralytics/.venv/lib/python3.12/site-packages/torch/nn/modules/container.py", line 250, in forward
    input = module(input)
  File "/root/ultralytics/ultralytics/nn/tasks.py", line 137, in forward
    return self.predict(x, *args, **kwargs)
  File "/root/ultralytics/ultralytics/nn/modules/head.py", line 359, in forward
    x = Detect.forward(self, x)
  File "/root/ultralytics/ultralytics/nn/modules/head.py", line 123, in forward
    y = self._inference(x)
Use tlparse to see full graph. (https://github.com/pytorch/tlparse?tab=readme-ov-file#tlparse-parse-structured-pt2-logs)

I exported the onnx with different method (labels.txt contains only "person"), I set the network_type=0 in the infer.txt. But I got this error
NvDsInferContext[UID 1]: Error in NvDsInferContextImpl::fillDetectionOutput() <nvdsinfer_context_impl_output_parsing.cpp:736> [UID = 1]: Failed to parse bboxes
./start.sh: line 3: 109256 Segmentation fault (core dumped) GST_DEBUG=3 python

With network_type=3 I got IndexError: list index out of range.
num_joints = int(obj_meta.mask_params.size / (sizeof(c_float) * 3)) is 2798

I sumply cant get it work with deeepstream.

[teddybee]

P.S: I think in my case the issue is with the ultralytics's internal onnx exporter. It doesn't do some mapping that the export_yolo script does (but fails with errors...). That's why I got so much points.
network_type=3 is still strange in the repo's infer.txt. I have read that yolo pose should be 0 or 100(custom).

P.S.2: The onnx export problem solved with downgrading/fixing the pytorch version to 2.8.0
The inference is good as well.