PyTorch_YOLOv4
PyTorch_YOLOv4 copied to clipboard
how should the data set be organized
Hello, I would like to train my data set, how should the data set be organized
/train.txt # list your path of training images
/val.txt # list your path of validation images
/images/
/train/
*.jpg
/val/
*.jpg
/labels/
/train/
*.txt # each line: category x_center_in_ratio y_center_in_ratio width_in_ratio height_in_ratio
/val/
*.txt # each line: category x_center_in_ratio y_center_in_ratio width_in_ratio height_in_ratio
/train.txt # list your path of training images /val.txt # list your path of validation images /images/ /train/ *.jpg /val/ *.jpg /labels/ /train/ *.txt # each line: category x_center_in_ratio y_center_in_ratio width_in_ratio height_in_ratio /val/ *.txt # each line: category x_center_in_ratio y_center_in_ratio width_in_ratio height_in_ratio
thank you
Can you give links to download the txt files labels (train2017 and val2017) for coco datasets? @WongKinYiu Any info on where to get data would be welcomed