Open nyj-ocean opened 5 years ago
Yes.
replace [yolo] by [Gaussian_yolo] in cfg-file and use filters = (classes + 8 + 1) * <numbers in mask>
in [convolutional] layers https://github.com/AlexeyAB/darknet/files/2853459/yolov3-spp_xnor_obj.cfg.txt
use this pre-trained weights-file https://drive.google.com/file/d/1d4CkgR--7bEEN0kWy-osR3kjLVFDIrnl/view?usp=sharing
Thanks for your great work! I want to combine Gaussian_YOLO with yolov3-spp_xnor_obj.cfg That means I want to use
[Gaussian_yolo]
instead of[yolo]
in the yolov3-spp_xnor_obj.cfg . and usefilters = (classes + 8 + 1) * <numbers in mask>
instead offilters = (classes + 4 + 1) * <numbers in mask>
in the yolov3-spp_xnor_obj.cfg . If this repository supports to train (Gaussian_YOLO+yolov3-spp_xnor_obj.cfg)? If supports,which pre-trained file should I use to train?