![]() ![]() We apply, for the first time, a deep learning method to automatically transcribe organ tablatures into modern music notation.The contributions of this paper are as follows: On the average, an error occurs every 220th pitch/rest character and every 833rd duration/special character. The neural network achieves an accuracy of 97.2% and 99.3% correctly recognized bars, depending on whether note pitch and rest characters or note duration and special characters are considered, respectively. We present the results of an experimental evaluation of the performance of the proposed approach. This data set and the tools to create it are available online. Using data augmentation and a synthetic data generator that we developed as part of our work, we generated a data set of sufficient size to perform the training. We utilize two scanned organ tablature books as data sources for training our deep neural network. The transcribed row consists of four tablature staves that are converted into a four-part score in modern notation. Transcription of a tablature row into modern music notation. An example of such an automatic transcription is shown in Figure 1. Then, the results of this process are converted to the format of Lilypond, 1 an open-source music notation program that can be used to generate a graphical output in modern notation. First, our method segments each input image into the corresponding tablature staves and recognizes tablature characters in the resulting partial images using a deep neural network. In this paper, we present a deep learning approach that automatically transcribes scanned organ tablature pages to modern music notation. ![]() Several archives contain large numbers of organ tablatures, some of which have neither been digitized nor been transcribed to modern notation yet ( Motnik, 2011 Wojnowska, 2016). It is studied by musicologists and is important to improve our knowledge about renaissance music. The New German Organ Tablature is one such old music notation. Manual transcription, however, is a time-consuming and error-prone process. Often, a manual transcription of the source material into modern music notation is required to make the material accessible to a wider audience and facilitate musicological analyses. The analysis of historical music notation is a major research topic in the field of musicology. Keywords: Organ Tablature, Automatic Transcription, Deep Learning, OCR, OMR DOI: Overall, our approach achieves an accuracy of 97.2% and 99.3% correctly recognized bars, depending on whether note pitch and rest characters or note duration and special characters are considered, respectively. We identify several types of error and validate that these are primarily caused by the poor legibility of relevant parts of some tablature scans. The results of our experiments are evaluated on tablatures taken from two tablature books. The artificial neural network model developed for the recognition of tablature characters is trained using a combination of real annotated tablature staves and tablatures produced by a synthetic data generator. Our approach is aimed at generating a uniform transcription that remains as close as possible to the original sheet music and therefore does not perform automatic error correction or musical interpretation. In this paper, we present a deep learning approach to automatically recognize organ tablature notation in scanned documents and transcribe it to modern music notation. The manual transcription of organ tablature compositions to modern music notation is time-consuming and often prone to errors. Organ tablature music notation differs considerably in structure and form from the music notation used today. Automatic Transcription of Organ Tablature Music Notation with Deep Neural Networks Research Automatic Transcription of Organ Tablature Music Notation with Deep Neural Networks Authors: Abstract
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |