yolov8-pose推理时报错

[FunnyWii]

funnywii@4060Ti:~/Documents/CV/PoseDet/TensorRT-Alpha/yolov8-pose/build$ ./app_yolov8_pose --model=../../data/yolov8/yolov8n-pose.trt --size=640 --batch_size=1 --img=../../data/6406407.jpg --show --savePath=../ [03/26/2024-16:29:24] [I] model_path = ../../data/yolov8/yolov8n-pose.trt [03/26/2024-16:29:24] [I] size = 640 [03/26/2024-16:29:24] [I] batch_size = 1 [03/26/2024-16:29:24] [I] image_path = ../../data/6406407.jpg [03/26/2024-16:29:24] [I] is_show = 0 [03/26/2024-16:29:24] [I] save_path = ../ [03/26/2024-16:29:24] [I] [TRT] Loaded engine size: 22 MiB [03/26/2024-16:29:24] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +18, now: CPU 0, GPU 18 (MiB) [03/26/2024-16:29:24] [I] [TRT] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +79, now: CPU 0, GPU 97 (MiB) [03/26/2024-16:29:24] [W] [TRT] CUDA lazy loading is not enabled. Enabling it can significantly reduce device memory usage and speed up TensorRT initialization. See "Lazy Loading" section of CUDA documentation https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#lazy-loading Segmentation fault (core dumped)

系统环境

Ubuntu 2004 CUDA 12.0 OpenCV 4.7.0 显卡 4060Ti

模型转换使用命令trtexec --onnx=yolov8n-pose.onnx --saveEngine=yolov8n-pose.trt --buildOnly --minShapes=images:1x3x640x640 --optShapes=images:2x3x640x640 --maxShapes=images:4x3x640x640

图片、视频、cam，推理均会在最后报错

[FeiYull]

@FunnyWii 需要debug确认哪一行报错