CPFelix / YOLOv5-LibTorch-Linux

Edit from https://github.com/Nebula4869/YOLOv5-LibTorch.

YOLOv5 LibTorch

Real time object detection with deployment of YOLOv5 through LibTorch C++ API With CUDA10.2.

Environment

• Ubuntu 16.04
• OpenCV 3.2.0
• LibTorch 1.7.0
• CMake 3.20.0
• Nvidia Driver 460.67
• CUDA10.2

Getting Started

1. Install OpenCV.

   sudo apt-get install libopencv-dev

   recommend using cmake to build opencv to specify the version.

2. Download LibTorch.

   https://blog.csdn.net/weixin_43742643/article/details/114156298

3. Edit "CMakeLists.txt" to configure OpenCV and LibTorch correctly.

4. Test model download

   truck detect model. 链接：https://pan.baidu.com/s/1PETg82SDpNb52Hjbh71BZQ 提取码：36ax

5. Compile and run.

   cd build
   cmake ..
   make
   ./../bin/YOLOv5LibTorch