Skip to main content

Evaluation of Lyrics Extraction from Folk Music Sheets Using Vision Language Models (VLMs)

  • Conference paper
  • First Online:
Progress in Artificial Intelligence (EPIA 2024)

Abstract

Monodic folk music has traditionally been preserved in physical documents. It constitutes a vast archive that needs to be digitized to facilitate comprehensive analysis using AI techniques. A critical component of music score digitization is the transcription of lyrics, an extensively researched process in Optical Character Recognition (OCR) and document layout analysis. These fields typically require the development of specific models that operate in several stages: first, to detect the bounding boxes of specific texts, then to identify the language, and finally, to recognize the characters. Recent advances in vision language models (VLMs) have introduced multimodal capabilities, such as processing images and text, which are competitive with traditional OCR methods. This paper proposes an end-to-end system for extracting lyrics from images of handwritten musical scores. We aim to evaluate the performance of two state-of-the-art VLMs to determine whether they can eliminate the need to develop specialized text recognition and OCR models for this task. The results of the study, obtained from a dataset in a real-world application environment, are presented along with promising new research directions in the field. This progress contributes to preserving cultural heritage and opens up new possibilities for global analysis and research in folk music.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

References

  1. ABBYY: ABBYY FineReader (2023). https://www.abbyy.com/en-eu/finereader. Accessed 25 May 2024

  2. Andrea, P., Zahra, A.: Music note position recognition in optical music recognition using convolutional neural network. Int. J. Arts Technol. 13(1), 45–60 (2021)

    Google Scholar 

  3. Audiveris: Audiveris omr (2021). https://audiveris.github.io/audiveris. Accessed 25 May 2024

  4. Biblioteca Nacional: Presentación del fondo de música tradicional imf-csic en madrid (2024). https://www.musicatradicional.eu/. Accessed 25 May 2024

  5. Cai, J., Peng, L., Tang, Y., Liu, C., Li, P.: Th-GAN: generative adversarial network based transfer learning for historical Chinese character recognition. In: 2019 International Conference on Document Analysis and Recognition (ICDAR), pp. 178–183 (2019). https://doi.org/10.1109/ICDAR.2019.00037

  6. Calvo-Zaragoza, J., Hajic, J., Pacha, A.: Understanding optical music recognition. ACM Comput. Surv. (CSUR) 53(4) (7 2020). https://doi.org/10.1145/3397499, https://dl.acm.org/doi/10.1145/3397499

  7. Castellanos, F.J., Calvo-Zaragoza, J., Inesta, J.M.: A neural approach for full-page optical music recognition of mensural documents. In: ISMIR, pp. 558–565 (2020)

    Google Scholar 

  8. Google Cloud: Google Cloud Vision OCR (2024). https://cloud.google.com/vision. Accessed 25 May 2024

  9. Hu, S., et al.: MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies (2024). https://arxiv.org/abs/2404.06395v3

  10. Huang, Y., Lv, T., Cui, L., Lu, Y., Wei, F.: Layoutlmv3: pre-training for document AI with unified text and image masking. In: Proceedings of the 30th ACM International Conference on Multimedia. MM ’22, pp. 4083–4091. Association for Computing Machinery, New York, NY, USA (2022). https://doi.org/10.1145/3503161.3548112

  11. Lee, S., Son, S.J., Oh, J., Kwak, N.: Handwritten music symbol classification using deep convolutional neural networks. In: 2016 International Conference on Information Science and Security (ICISS), pp. 1–5. IEEE (2016)

    Google Scholar 

  12. Liu, Y., et al.: On the hidden mystery of OCR in large multimodal models (2023). https://arxiv.org/abs/2305.07895v5

  13. Mendes, A.S., Lozano Murciego, L., Silva, L.A., Jiménez Bravo, D.M., Navarro Cáceres, M., Bernardes, G.: salesmendesandre/dataset-Evaluation-of-Lyrics- Extraction-from-Folk-Music-Sheets-Using-Vision- Language-Model: Dataset Release v1 (2024). https://doi.org/10.5281/zenodo.12662916

  14. OpenAI: OpenAI API (2024). https://platform.openai.com/docs/models/gpt-4o. Accessed 03 July 2024

  15. Saluja, R., Punjabi, M., Carman, M., Ramakrishnan, G., Chaudhuri, P.: Sub-word embeddings for OCR corrections in highly fusional Indic languages. In: 2019 International Conference on Document Analysis and Recognition (ICDAR), pp. 160–165 (2019). https://doi.org/10.1109/ICDAR.2019.00034

  16. Smith, R.: An overview of the tesseract OCR engine. In: Ninth International Conference on Document Analysis and Recognition (ICDAR 2007), vol. 2, pp. 629–633 (2007). https://doi.org/10.1109/ICDAR.2007.4376991

  17. Tensmeyer, C., Davis, B., Wigington, C., Lee, I., Barrett, B.: Pagenet: page boundary extraction in historical handwritten documents. In: Proceedings of the 4th International Workshop on Historical Document Imaging and Processing, pp. 59–64 (2017)

    Google Scholar 

  18. Wang, W., et al.: Cogvlm: visual expert for pretrained language models (2023)

    Google Scholar 

  19. van der Wel, E., Ullrich, K.: Optical music recognition with convolutional sequence-to-sequence models. In: Proceedings of the 18th International Society for Music Information Retrieval Conference, pp. 731–737. ISMIR (2018). https://doi.org/10.5281/zenodo.1415664

Download references

Acknowledgments

We would like to express our gratitude for the assistance from Gilberto Bernardes and Nádia Carvalho from INESC TEC - Institute for Systems and Computer Engineering, Technology and Science in carrying out the case study of this research. Research funded by the project “EA-DIGIFOLK: AN EUROPEAN AND IBERO-AMERICAN APPROACH FOR THE DIGITAL COLLECTION, ANALYSIS AND DISSEMINATION OF FOLK MUSIC” (101086338) under the program Marie Skłodowska-Curie Actions Staff Exchanges (HORIZON-MSCA-2021-SE-01-01).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to André Sales Mendes .

Editor information

Editors and Affiliations

Ethics declarations

Disclosure of Interests

The authors have no competing interests to declare relevant to this article’s content.

Rights and permissions

Reprints and permissions

Copyright information

© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Sales Mendes, A., Lozano Murciego, Á., Silva, L.A., Jiménez-Bravo, D.M., Navarro-Cáceres, M., Bernardes, G. (2025). Evaluation of Lyrics Extraction from Folk Music Sheets Using Vision Language Models (VLMs). In: Santos, M.F., Machado, J., Novais, P., Cortez, P., Moreira, P.M. (eds) Progress in Artificial Intelligence. EPIA 2024. Lecture Notes in Computer Science(), vol 14967. Springer, Cham. https://doi.org/10.1007/978-3-031-73497-7_8

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-73497-7_8

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-73496-0

  • Online ISBN: 978-3-031-73497-7

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics