Original codes from tensorrtx. I modified the yololayer and integrated batchedNMSPlugin. A yolov5s.wts is provided for fast demo. How to generate .wts can refer to https://github.com/wang-xinyu/tensorrtx/tree/master/yolov5.
- build and run
mkdir build
cd build
cmake ..
make
sudo ./yolov5 -s // serialize model to plan file i.e. 'yolov5s.engine'
sudo ./yolov5 -d ../samples // deserialize plan file and run inference, the images in samples will be processed.
- check the images generated, as follows. _zidane.jpg and _bus.jpg
- run Python example, please install Python tensorrt and Pycuda and then
python yolov5_trt.py
See the readme in tensorrtx home page.
None!