Dataset · PRAIG
Sheet music recognition (OMR) benchmark
PRAIG/SMB
Printed sheet music images with region annotations for testing optical music recognition systems.
What's inside
This dataset contains up to 1,000 examples of printed scores in Common Western Modern Notation, structured as image-text pairs. Collected by the Pattern Recognition and Artificial Intelligence Group at the University of Alicante, with file access granted upon request.
What you can do with it
- Evaluating full-page optical music recognition (OMR) systems
- Training image segmentation models to isolate musical regions on a page
- Testing image-to-text multimodal models on sheet music reading
- Tasks
- Image → text, Image segmentation, Text retrieval
- Size
- up to 1K rows
- Downloads this month
- 2.5K
- Likes
- 17
- Access
- on request on Hugging Face
How to load it
from datasets import load_dataset
ds = load_dataset("PRAIG/SMB")