Training dataset for the Tzefa YOLO11x-OBB line segmentation model.
data.yaml
train/
images/ # Training images (.png + optional .npy)
labels/ # OBB annotations (.txt)
val/
images/ # Validation images
labels/ # Validation annotations
YOLO OBB format: class x1 y1 x2 y2 x3 y3 x4 y4 (normalized coordinates)
Each label file contains one line per text line detected… See the full description on the dataset page:
https://huggingface.co/datasets/WARAJA/Tzefa-Line-Segmentation-Dataset.