search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.71k stars 7.96k forks source link

Custom Model Architecture Using DarkNet #8067

Open RAJA-PARIKSHAT opened 3 years ago

RAJA-PARIKSHAT commented 3 years ago

I trained a custom face mask detection model using tiny yolov4 but I didn't make any change in the architecture, let's suppose I want to modify architecture can I do so, and how? Its like I don't want to use pretrained model rather want to train on custom model from scratch

lsd1994 commented 3 years ago

Yes you can, you should understand the meaning of each layer first and design your own architecture. https://github.com/AlexeyAB/darknet/wiki/CFG-Parameters-in-the-different-layers