Run detections on EfficientDet and YOLOv3 models with onnx.
ONNX Tutorials
EfficientDet implementation based from: https://github.com/murdockhou/Yet-Another-EfficientDet-Pytorch-Convert-ONNX-TVM
YOLOv3 implementation based from: https://github.com/AlexeyAB/darknet
Use the package manager pip.
pip install requirements.txt.onnx model
/onnx_runtime/models/myOnnxModel.onnx
test images
/onnx_runtime/data/images
python detect.py <original framework> <path to onnx model> <path to test images>
Output:
Inference time: 0.18 sec
3 objects identified in source image.
Credit to the tutorials and documents provided by onnx.
https://github.com/onnx
