AlexeyAB
commented
6 years ago @Jin-S-Kim Hi,
You can't merge two different models into one.
You should create your own dataset, that includes all required labeled objects and train your own model. To get MSCOCO dataset (with labeled objects) - use this script on Linux: https://github.com/AlexeyAB/darknet/blob/master/scripts/get_coco_dataset.sh As described here: https://pjreddie.com/darknet/yolo/
If MSCOCO downloads slowly, then you can download it from mscoco-site:
- http://images.cocodataset.org/zips/train2014.zip
- http://images.cocodataset.org/zips/val2014.zip
- http://images.cocodataset.org/annotations/annotations_trainval2014.zip
Then comment these lines: https://github.com/AlexeyAB/darknet/blob/18d5e4f39c1441f2c21043ac3204b5cb279f8758/scripts/get_coco_dataset.sh#L11-L12
And the run get_coco_dataset.sh
Hello, I want to train ingredient by using Yolo_mark. When I downloaded darknet, some ingredients are already trained like this . I can check these names file name of coco.names
And i want to detect object together but I can't. I train zucchini and I want to detect zucchini and apple together but weight files that I trained and I downloaded is different. So I can't detect apple and zucchini together.
As I know it, yolov3.weights files are already trained by Yolo_mark. Can I downloaded image files that is used when making yolov3.weights?? If I can't, How can I detect objects that are in coco.names and zucchini together by training only zucchini?