zzsi/deep-scores-v2-dense

DeepScoresV2 — Dense Subset

A HuggingFace-formatted mirror of the dense subset of the DeepScoresV2 dataset for music object detection.

Dataset description

DeepScoresV2 is a large-scale dataset of synthetically rendered music score pages annotated with bounding boxes for musical symbols. The dense subset contains 1,714 images selected by the authors as the most diverse and representative sample from the full 803k-image dataset.

Each image is a full score page rendered from MuseScore. Annotations follow COCO format: bbox is [x, y, width, height] in pixel coordinates.

Format

{
    "image_id":  int,
    "file_name": str,
    "image":     PIL.Image,     # full score page
    "width":     int,
    "height":    int,
    "objects": {
        "id":          List[int],
        "bbox":        List[List[float]],  # [x, y, w, h], COCO format
        "category_id": List[int],
        "category":    List[str],          # symbol class name
        "area":        List[float],
        "iscrowd":     List[int],
    },
}

Usage

from datasets import load_dataset

ds = load_dataset("zzsi/deep-scores-v2-dense")
example = ds["train"][0]
print(example["objects"]["category"][:5])
example["image"].show()

License

Creative Commons Attribution 4.0 International (CC BY 4.0)

Attribution

This dataset is a reformatted mirror of DeepScoresV2. Please cite the original work:

@inproceedings{DeepScoresV2,
  title     = {DeepScoresV2: A Dataset for Music Object Detection with a Challenging Test Set},
  author    = {Tuggener, Lukas and Satyawan, Yvan Putra and Pacha, Alexander
               and Schmidhuber, J{\"u}rgen and Stadelmann, Thilo},
  booktitle = {British Machine Vision Conference (BMVC)},
  year      = {2021}
}

Original dataset: https://zenodo.org/records/4012193 Original authors: Lukas Tuggener, Yvan Putra Satyawan, Alexander Pacha, Jürgen Schmidhuber, Thilo Stadelmann (ZHAW / IDSIA)