guangyangmusic commited on
Commit
7a11fe0
·
verified ·
1 Parent(s): 2aeb72c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +110 -24
README.md CHANGED
@@ -1,24 +1,110 @@
1
- ---
2
- license: cc-by-4.0
3
- dataset_info:
4
- features:
5
- - name: image_imslp
6
- dtype: image
7
- - name: image_mscore
8
- dtype: image
9
- - name: musicxml
10
- dtype: string
11
- - name: filename
12
- dtype: string
13
- splits:
14
- - name: train
15
- num_bytes: 287742576.0
16
- num_examples: 252
17
- download_size: 245568520
18
- dataset_size: 287742576.0
19
- configs:
20
- - config_name: default
21
- data_files:
22
- - split: train
23
- path: data/train-*
24
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ task_categories:
4
+ - image-to-text
5
+ language:
6
+ - en
7
+ tags:
8
+ - music
9
+ - optical-music-recognition
10
+ - OMR
11
+ - MusicXML
12
+ - sheet-music
13
+ pretty_name: OpenScore String Quartets (OMR Evaluation)
14
+ size_categories:
15
+ - n<1K
16
+ dataset_info:
17
+ features:
18
+ - name: image_imslp
19
+ dtype: image
20
+ - name: image_mscore
21
+ dtype: image
22
+ - name: musicxml
23
+ dtype: string
24
+ - name: filename
25
+ dtype: string
26
+ splits:
27
+ - name: train
28
+ num_examples: 252
29
+ ---
30
+
31
+ # OpenScore String Quartets (OMR Evaluation)
32
+
33
+ This dataset is derived from the [OpenScore String Quartets](https://github.com/OpenScore/StringQuartets) corpus (Gotham et al., 2023), a collection of string quartets by "long 19th century" composers. It is designed for evaluating Optical Music Recognition (OMR) systems.
34
+
35
+ We extract a subset of the OpenScore String Quartets that contains both scanned images of real scores and the corresponding MusicXML ground truth. We also render clean images from the MusicXML files using MuseScore. This gives two evaluation settings:
36
+
37
+ - **Camera** (`image_imslp`): scanned page images from [IMSLP](https://imslp.org/), reflecting real-world conditions (noise, aging, varying scan quality).
38
+ - **Rendered** (`image_mscore`): clean page images rendered from the ground-truth MusicXML via MuseScore.
39
+
40
+ ## Dataset Structure
41
+
42
+ | Column | Type | Description |
43
+ |---|---|---|
44
+ | `image_imslp` | `Image` | Scanned page image from IMSLP (camera) |
45
+ | `image_mscore` | `Image` | Clean rendered page image from MuseScore |
46
+ | `musicxml` | `string` | Ground-truth MusicXML content |
47
+ | `filename` | `string` | Composition and page identifier |
48
+
49
+ The dataset contains **252** page-level examples.
50
+
51
+ ## Usage
52
+
53
+ ```python
54
+ from datasets import load_dataset
55
+
56
+ ds = load_dataset("guangyangmusic/OpenScore-StringQuartets", split="train")
57
+
58
+ # access an example
59
+ example = ds[0]
60
+ example["image_imslp"].show() # scanned image
61
+ example["image_mscore"].show() # rendered image
62
+ print(example["musicxml"][:200])
63
+ print(example["filename"])
64
+ ```
65
+
66
+ ## Source
67
+
68
+ The original OpenScore String Quartets corpus is available at [https://github.com/OpenScore/StringQuartets](https://github.com/OpenScore/StringQuartets) and on [MuseScore](https://musescore.com/openscore-string-quartets).
69
+
70
+ ## License
71
+
72
+ The underlying [OpenScore String Quartets](https://github.com/OpenScore/StringQuartets) scores are in the public domain under [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/).
73
+
74
+ This repository is a **derived** resource (subset selection, IMSLP page images, MuseScore renders, aligned MusicXML strings, and packaging). We release it under [Creative Commons Attribution 4.0 International (CC BY 4.0)](https://creativecommons.org/licenses/by/4.0/). You may share and adapt it if you give **appropriate credit**, including **this dataset** and the **original OpenScore corpus** (see citation below).
75
+
76
+ ## Citation
77
+
78
+ If you use this dataset, please cite **both** this release (Legato evaluation subset) and the original OpenScore String Quartets corpus.
79
+
80
+ **This dataset / Legato paper:**
81
+
82
+ ```bibtex
83
+ @misc{yang2025legatolargescaleendtoendgeneralizable,
84
+ title={LEGATO: Large-scale End-to-end Generalizable Approach to Typeset OMR},
85
+ author={Guang Yang and Victoria Ebert and Nazif Tamer and Brian Siyuan Zheng and Luiza Pozzobon and Noah A. Smith},
86
+ year={2025},
87
+ eprint={2506.19065},
88
+ archivePrefix={arXiv},
89
+ primaryClass={cs.CV},
90
+ url={https://arxiv.org/abs/2506.19065},
91
+ }
92
+ ```
93
+
94
+ **Original OpenScore String Quartets corpus:**
95
+
96
+ ```bibtex
97
+ @inproceedings{gotham_openscore_2023,
98
+ address = {Milan, Italy},
99
+ title = {The "{OpenScore} {String} {Quartet}" {Corpus}},
100
+ isbn = {9798400708336},
101
+ url = {https://dl.acm.org/doi/10.1145/3625135.3625155},
102
+ doi = {10.1145/3625135.3625155},
103
+ booktitle = {Proceedings of the 10th International Conference on Digital Libraries for Musicology},
104
+ publisher = {ACM},
105
+ author = {Gotham, Mark R. H. and Redbond, Maureen and Bower, Bruno and Jonas, Peter},
106
+ month = nov,
107
+ year = {2023},
108
+ pages = {49--57},
109
+ }
110
+ ```