- Easily implement producer-consumer models for various tasks and perform high-performance inference
- No complex packaging, no coupling!
- Currently supports Yolo series 3/4/5/x/7/8
- YoloV8-Segment is supported
- 🚀 Pre-processing about 1ms
- 🚀 Post-processing about 0.5ms
- cpm.hpp Producer-consumer model
- For direct inference tasks, cpm.hpp can be turned into an automatic multi-batch producer-consumer model
- infer.hpp A repackaging of tensorRT. Simple interface
- yolo.hpp Wrapper for yolo tasks. Based on infer.hpp
trtexec --onnx=yolov5s.onnx --saveEngine=yolov5s.engine
model = trt::load("yolov5s.engine");
... preprocess ...
// Configure the dynamic batch size.
auto dims = model->static_dims();
dims[0] = batch;
model->set_run_dims(dims);
model->forward({input_device, output_device}, stream);
... postprocess ...cv::Mat image = cv::imread("image.jpg");
auto model = yolo::load("yolov5s.engine");
auto objs = model->forward(yolo::Image(image.data, image.cols, image.rows));
// use objs to draw to image. cpm::Instance<yolo::BoxArray, yolo::Image, yolo::Infer> cpmi;
cpmi.start([]{
return yolo::load("yolov5s.engine", yolo::Type::V5);
}, batch);
auto result_futures = cpmi.commits(images);
for(auto& fut : result_futures){
auto objs = fut.get();
... process ...
}
