I'm trying to train a segmentation model using the FracAtlas dataset, but I'm encountering an error indicating that the dataset isn't correctly formatted for this type of task/model.
I'm splitting my data using the utility notebook (FracAtlas/notebooks/dataset/Utilities/coco2yolo.ipynb) provided for conversion of COCO annotation to YOLO format, and then splitting of images and labels as it's needed. While this works wonderfully for object detection, it doesn't seem to work for segmentation.
Any advice or assistance would be greatly appreciated!
Hi, everyone!
I'm trying to train a segmentation model using the FracAtlas dataset, but I'm encountering an error indicating that the dataset isn't correctly formatted for this type of task/model.
I'm splitting my data using the utility notebook (FracAtlas/notebooks/dataset/Utilities/coco2yolo.ipynb) provided for conversion of COCO annotation to YOLO format, and then splitting of images and labels as it's needed. While this works wonderfully for object detection, it doesn't seem to work for segmentation.
Any advice or assistance would be greatly appreciated!
Thanks in advance.