I am not sure how the custom object detection training should be done similar to yoloV3 or yolov4 using darknet framework and darknet weights, but with CSP backbones.
I would like to use the CSP backbones for example,
CSPResNeXt50-PANet-SPP | 608×608 | with AP50 = 65.4 | and has corresponding cfg and weight files to download.
Is the procedure same as with the yolo with darknet weights? Please write me the steps or the approach to use CSP backbones.
Hi,
I am not sure how the custom object detection training should be done similar to yoloV3 or yolov4 using darknet framework and darknet weights, but with CSP backbones.
I would like to use the CSP backbones for example, CSPResNeXt50-PANet-SPP | 608×608 | with AP50 = 65.4 | and has corresponding cfg and weight files to download.
Is the procedure same as with the yolo with darknet weights? Please write me the steps or the approach to use CSP backbones.
Thanks