NOTE: The yaml file is not required.
git clone https://github.com/derronqi/yolov7-face.git
cd yolov7-face
pip3 install -r requirements.txt
pip3 install onnx onnxsim onnxruntime
NOTE: It is recommended to use Python virtualenv.
Copy the export_yoloV7_face.py
file from DeepStream-Yolo-Face/utils
directory to the yolov7-face
folder.
Download the pt
file from YOLOv7-Face repo.
NOTE: You can use your custom model.
Generate the ONNX model file (example for YOLOv7-Face)
python3 export_yoloV7_face.py -w yolov7-face.pt --dynamic
NOTE: To change the inference size (defaut: 640)
-s SIZE
--size SIZE
-s HEIGHT WIDTH
--size HEIGHT WIDTH
Example for 1280
-s 1280
or
-s 1280 1280
NOTE: To simplify the ONNX model (DeepStream >= 6.0)
--simplify
NOTE: To use dynamic batch-size (DeepStream >= 6.1)
--dynamic
NOTE: To use static batch-size (example for batch-size = 4)
--batch 4
Copy the generated ONNX model file to the DeepStream-Yolo-Face
folder.
Edit the config_infer_primary_yoloV7_face.txt
file according to your model (example for YOLOv7-Face)
[property]
...
onnx-file=yolov7-face.onnx
...
NOTE: The YOLOv7-Face resizes the input with center padding. To get better accuracy, use
[property]
...
maintain-aspect-ratio=1
symmetric-padding=1
...