- Clone this repository
$ git clone https://github.com/tomerni/bayBwatch- Download the weights files:
** yolov3-tiny weights** https://drive.google.com/file/d/1_XixOaRpi26B-3ykxQBBwp02Uk5CZ5O7/view?usp=sharing
** yolov3-face weights** https://drive.google.com/file/d/1YFOxAtOH6Xa-1VuIJIdRsWAZxsj6rhuO/view?usp=sharing
- Install requirements:
$ pip install -r requirements.txt- Run the function _main() in yoloface_api.py:
def _main(model_cfg: str, model_weights: str, output_dir: str, image="", video="", src=None):- Supply the relevant config and weights (See CLI usage for more information about the detection types).
- Must supply image path/video path/src=1(webcam)
- For body detection run the following command (change the IMG_PATH/VIDEO_PATH):
image input
$ python yoloface.py --model-cfg "./cfg/yolov3-tiny.cfg" --model-model-model-weights "./model-weights/yolov3-tiny.weights" --image IMG_PATH --output-dir outputs/video input
$ python yoloface.py --model-cfg "./cfg/yolov3-tiny.cfg" --model-model-model-weights "./model-weights/yolov3-tiny.weights" --image VIDEO_PATH --output-dir outputs/webcam
$ python yoloface.py --model-cfg "./cfg/yolov3-tiny.cfg" --model-model-model-weights "./model-weights/yolov3-tiny.weights" --src 0 --output-dir outputs/- For face detection run the following command:
image input
$ python yoloface.py --model-cfg "./cfg/yolov3-face.cfg" --model-model-model-weights "./model-weights/yolov3-wider_16000.weights" --image IMG_PATH --output-dir outputs/video input
$ python yoloface.py --model-cfg "./cfg/yolov3-face.cfg" --model-model-model-weights "./model-weights/yolov3-wider_16000.weights" --image VIDEO_PATH --output-dir outputs/webcam
$ python yoloface.py --model-cfg "./cfg/yolov3-face.cfg" --model-model-model-weights "./model-weights/yolov3-wider_16000.weights" --src 0 --output-dir outputs/- --model-cfg - yolov3 config path
- --model-weights - yolov3 weights path
- --image/video - image/video path
- --src - 0 for webcam
- --output-dir - the output's path