Search

Start Over Contributor Grantová agentura Univerzity Karlovy v Praze@@GAUK 289623@@Generování syntetických trénovacích dat a jiné metody pro rozpoznávání psaných notopisů@@nationalFunds@@ Language No linguistic content Harvested from LINDAT/CLARIAH-CZ repository

1. GrandStaff-LMX: Linearized MusicXML Encoding of the GrandStaff Dataset

Creator:
Mayer, Jiří, Straka, Milan, Hajič jr., Jan, and Pecina, Pavel
Publisher:
Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics (UFAL)
Type:
image and corpus
Subject:
GrandStaff, pianoform scores, MusicXML, and Linearized MusicXML
Language:
No linguistic content
Description:
The GrandStaff-LMX dataset is based on the GrandStaff dataset described in the "End-to-end optical music recognition for pianoform sheet music" paper by Antonio Ríos-Vila et al., 2023, https://doi.org/10.1007/s10032-023-00432-z . The GrandStaff-LMX dataset contains MusicXML and Linearized MusicXML encodings of all systems from the original datase, suitable for evaluation with the TEDn metric. It also contains the GrandStaff official train/dev/split.
Rights:
Creative Commons - Attribution-ShareAlike 4.0 International (CC BY-SA 4.0), http://creativecommons.org/licenses/by-sa/4.0/, and PUB

2. OLiMPiC 1.0: OpenScore Lieder Linearized MusicXML Piano Corpus

Creator:
Mayer, Jiří, Straka, Milan, Hajič jr., Jan, and Pecina, Pavel
Publisher:
Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics (UFAL)
Type:
image and corpus
Subject:
OpenScore Lieder, pianoform scores, MusicXML, and Linearized MusicXML
Language:
No linguistic content
Description:
OLiMPiC: OpenScore Lieder Linearized MusicXML Piano Corpus is a dataset containing synthetic and scanned images of pianoform music scores. The scores and the scanned images originate from the OpenScore Lieder Corpus https://github.com/OpenScore/Lieder . OLiMPiC contains the scores in MusicXML and Linearized MusicXML encoding, suitable for evaluation with the TEDn metric. The official train/dev/test split is also provided.
Rights:
Creative Commons - Attribution-ShareAlike 4.0 International (CC BY-SA 4.0), http://creativecommons.org/licenses/by-sa/4.0/, and PUB