Repository files navigation

news: 2021.10.31:yolov5-v6.0 support

The project is the encapsulation of nvidia official yolo-tensorrtimplementation. And you must have the trained yolo model(.weights) and.cfg file from the darknet (yolov3 & yolov4). For theyolov5 ,you should prepare the model file (yolov5s.yaml) and the trained weight file (yolov5s.pt) from pytorch.

• yolov5n ,yolov5s , yolov5m , yolov5l , yolov5x ,yolov5-p6tutorial
• yolov4
• yolov3

• inequal net width and height
• batch inference
• support FP32,FP16,INT8
• dynamic input size

• windows 10
• ubuntu 18.04
• L4T (Jetson platform)

Prepare the pretrained.weights and.cfg model.

Detector detector;Config config;std::vector<BatchResult> res;detector.detect(vec_image, res)

• dependency : TensorRT 7.1.3.4 , cuda 11.0 , cudnn 8.0 , opencv4 , vs2015

• build:

  open MSVCsln/sln.sln file

  • dll project : the trt yolo detector dll
  • demo project : test of the dll

The project generate thelibdetector.so lib, and the sample code.If you want to use the libdetector.so lib in your own project,thiscmake file perhaps could help you .

git clone https://github.com/enazoe/yolo-tensorrt.gitcd yolo-tensorrt/mkdir buildcd build/cmake ..make./yolo-trt

structConfig{std::string file_model_cfg ="configs/yolov4.cfg";std::string file_model_weights ="configs/yolov4.weights";float detect_thresh =0.9;ModelType net_type = YOLOV4;Precision inference_precison = INT8;int gpu_id =0;std::string calibration_image_list_file_txt ="configs/calibration_images.txt";};classAPI Detector{public:explicitDetector();~Detector();voidinit(const Config &config);voiddetect(const std::vector<cv::Mat> &mat_image,std::vector<BatchResult> &vec_batch_result);private:Detector(const Detector &);const Detector &operator =(const Detector &);classImpl;Impl *_impl;};

• https://github.com/wang-xinyu/tensorrtx/tree/master/yolov4
• https://github.com/mj8ac/trt-yolo-app_win64
• https://github.com/NVIDIA-AI-IOT/deepstream_reference_apps

微信关注公众号EigenVison，回复yolo获取交流群号