s94285 / onnx2darknet

onnx2darknet

Convert yolo2 model of ONNX format to darknet format

Get ONNX format

• https://github.com/onnx/models/tree/master/vision/object_detection_segmentation/tiny_yolov2

Convert ONNX format -> Darknet format

• python onnx2darknet.py 'yolov2.onnx' 'yolov2.cfg' 'yolov2.weights'

Inference Test(Darknet framework)

try predict

• Darknet Framework installation

  git clone https://github.com/pjreddie/darknet
  cd darknet
  make 

• Run detector
  • ./darknet detect cfg/yolov2.cfg yolov2.weights data/dog.jpg

Currently supported

Models

• Yolov2-tiny
• Yolov2(partial)

Operator

• BatchNormalization
• Conv
• LeakyRelu
• MaxPool

Requirements and Develop Environment

• python >= 2.7
• onnx >= 1.2.1
• darknet

Reference

• Using ONNX for Inference (YOLO2): [https://github.com/purelyvivid/yolo2_onnx]
• Detection Using A Pre-Trained Model: [https://pjreddie.com/darknet/yolo/]