This simple PR fixes the default image extension choices when using label-studio-converter to import YOLO data.
In yolo.py, the function definition for convert_yolo_to_ls uses image_ext=".jpg,.jpeg,.png" as the default image extension options. However, this default choice is overridden by the default --image-ext argument, which is just ".jpg".
The PR changes the default choices in the --image-ext argument to match the function default, because it seems like that's what was intended by the original authors of the code.
It changes the default --image-ext argument from ".jpg" to ".jpg,.jpeg.png"
Ultimately, this change makes it easier for a new user to use label-studio-converter without having to know about the --image-ext argument, which isn't documented anywhere. The Tutorial: Importing Local YOLO Pre-Annotated Images to Label Studio doesn't discuss the --image_ext argument at all, so new users could potentially be tripped up when they try to import images that aren't .jpg.
This simple PR fixes the default image extension choices when using label-studio-converter to import YOLO data.
yolo.py, the function definition forconvert_yolo_to_lsusesimage_ext=".jpg,.jpeg,.png"as the default image extension options. However, this default choice is overridden by the default--image-extargument, which is just".jpg".--image-extargument to match the function default, because it seems like that's what was intended by the original authors of the code.--image-extargument from".jpg"to".jpg,.jpeg.png"Ultimately, this change makes it easier for a new user to use
label-studio-converterwithout having to know about the--image-extargument, which isn't documented anywhere. The Tutorial: Importing Local YOLO Pre-Annotated Images to Label Studio doesn't discuss the--image_extargument at all, so new users could potentially be tripped up when they try to import images that aren't.jpg.