Redirigiendo al acceso original de articulo en 24 segundos...
Inicio  /  Algorithms  /  Vol: 15 Par: 7 (2022)  /  Artículo
ARTÍCULO
TITULO

Optical Medieval Music Recognition Using Background Knowledge

Alexander Hartelt and Frank Puppe    

Resumen

This paper deals with the effect of exploiting background knowledge for improving an OMR (Optical Music Recognition) deep learning pipeline for transcribing medieval, monophonic, handwritten music from the 12th?14th century, whose usage has been neglected in the literature. Various types of background knowledge about overlapping notes and text, clefs, graphical connections (neumes) and their implications on the position in staff of the notes were used and evaluated. Moreover, the effect of different encoder/decoder architectures and of different datasets for training a mixed model and for document-specific fine-tuning based on an extended OMR pipeline with an additional post-processing step were evaluated. The use of background models improves all metrics and in particular the melody accuracy rate (mAR), which is based on the insert, delete and replace operations necessary to convert the generated melody into the correct melody. When using a mixed model and evaluating on a different dataset, our best model achieves without fine-tuning and without post-processing a mAR of 90.4%, which is raised by nearly 30% to 93.2% mAR using background knowledge. With additional fine-tuning, the contribution of post-processing is even greater: the basic mAR of 90.5% is raised by more than 50% to 95.8% mAR.

 Artículos similares

       
 
Luca M. Ofiera, Purnendu Bose and Christian Kazner    
Constructed wetlands are a versatile technology for various treatment approaches, especially in emerging countries. The research aims to study and optimize the hybridizing process of a vertical subsurface flow constructed wetland with adsorption technolo... ver más
Revista: Water

 
Youngkwang Kim, Woochan Kim, Jungwoo Yoon, Sangkug Chung and Daegeun Kim    
This paper presents a practical contamination detection system for camera lenses using image analysis with deep learning. The proposed system can detect contamination in camera digital images through contamination learning utilizing deep learning, and it... ver más
Revista: Information

 
Bofu Zheng, Dan Wang, Yuxin Chen, Yihui Jiang, Fangqing Hu, Liliang Xu, Jihong Zhang and Jinqi Zhu    
Background: Vegetation roots are considered to play an effective role in controlling soil erosion by benefiting soil hydrology and mechanical properties. However, the correlation between soil hydrology and the mechanical features associated with the vari... ver más
Revista: Water

 
Gholamreza Eslamifar, Hamid Balali and Alexander Fernald    
Enhancing the comprehension of alterations in land use holds paramount importance for water management in semi-arid regions due to its effects on hydrology and agricultural economics. Allowing agricultural land to lie fallow has emerged as a technique to... ver más
Revista: Water

 
Noelia García, Miguel Gómez-Polo, Miriam Fernández, José Luis Antonaya-Martín, Rocío Ortega, Cristina Gómez-Polo, Marta Revilla-León and Rocío Cascos    
Background: With CAD-CAM, dental models are often fabricated by additive manufacturing (AM) methods. Dental models for implant prostheses involve the manual placement of analogs, which could affect their final fit. Therefore, in this experimental in vitr... ver más
Revista: Applied Sciences