Use pytorch1.5 to train yolov5 and export the onnx model. Use onnxruntime to run the onnx model on the cpu. yolov5 refers to https://github.com/ultralytics/yolov5
-
Face detection model: yolov5
-
requirement:
- opencv-python
- numpy
- onnxruntime-1.5.2 (cpu)
-
onnx model: download from https://drive.google.com/file/d/1jCT7-dpeHhCJ0DBrx4e1Vap9EFaguZs3/view?usp=sharing -
run:
python onnx_infer.py --onnx_path=./best.onnx --data_dir=./input