Optical Medieval Music Recognition Using Background Knowledge
Round 1
Reviewer 1 Report
The paper is a very interest application of a modern and actual tech to a field needing attention. The work has the potential to improve the current knowledge toward saving many man hour time and preserve human culture.
Avoid the use of we in English scientific writing style.
Remove pictures from the introduction, and place it somewhere else.
In the conclusions, please better summarize the work done (focus on methodology followed) and also report the issues occurred, the issues solved and the issues not solved. This will help other follow the work and support the scientific community.
Author Response
Dear Reviewer/Editor,
We appreciate the time and effort that you and the reviewers have dedicated to providing your valuable feedback on our manuscript. We are grateful for their insightful comments on our
paper. We have been able to incorporate changes to reflect most of the suggestions provided
by the reviewers. We have highlighted the changes within the manuscript. Here is a point-by-point response to the reviewers’ comments and concerns.
Comment 1: Avoid the use of we in English scientific writing style.
Response: Thank you for pointing this out. We agree with this comment.
Therefore, we have adjusted this in most places.
Location in manuscript: Since it runs through the entire document, we do not give the exact line number of the changes here.
Comment 2: Remove pictures from the introduction, and place it somewhere else.
Response: Thank you for pointing this out. We agree with this comment.
Therefore, we moved the image of the sample staff from the introduction to the dataset section. In addition, the image has been slightly expanded so that examples for the classes are annotated in the image. The caption has also been expanded to provide information about the origin of the image.
Location in manuscript: Figure 1
Comment 3: In the conclusions, please better summarize the work done (focus on methodology followed) and also report the issues occurred, the issues solved and the issues not solved. This will help other follow the work and support the scientific community.
Response: Thank you for pointing this out. We agree with this
comment.
Therefore, we stated which types of background knowledge were used and evaluated. We added a paragraph about further possible post-processing steps, which can be addressed in future works.
in addition, we gave another perspective on the use of background knowledge at the end.
Location in manuscript: Around Line 514 – 518, Line 531 – 533, Line 539 - 542
Additional clarifications
We added a the Acknowledgments section:
Location in manuscript: Around Line 548 - 549
We changed the Funding section:
Location in manuscript: Around Line 546 - 547
Reviewer 2 Report
Optical Music Recognition (OMR) is one of the key technologies to accelerate and simplify the task digitization of historical musical documents written by hand in an automatic way. Historical documents are written using a special notation, and general purpose tools are not suitable for this. In addition, a significant problem is the presence of artifacts in such recordings, such as continuous notes or a lost musical key, which requires the use of hidden knowledge to restore the melody. Authors consider scanning with Fully Convolutional Network (FCN) of U-Net type. The full training dataset cosists of about 230 pages of sheet music of 12-14 century manuscripts written in squared notation.
Authors main contribution is that they formalize parts of the expert knowledge and incorporate it into a post-processing pipeline to improve the symbol detection accuracy. The schematic workflow of the proposed symbol detection is presented. Post-processing includes 6 steps and allows correcting common errors of symbol detection.
I suppose that the manuscript presents a soild well-written paper of a high scientific level.
As a drawback, I may recommend formalizing what clearly background knowlegde is.
Author Response
Dear Reviewer/Editor,
We appreciate the time and effort that you and the reviewers have dedicated to providing your valuable feedback on our manuscript. We are grateful for their insightful comments on our
paper. We have been able to incorporate changes to reflect most of the suggestions provided
by the reviewers. We have highlighted the changes within the manuscript. Here is a point-by-point response to the reviewers’ comments and concerns.
Comment 1: As a drawback, I may recommend formalizing what clearly background knowlegde is.
Response: Thank you for pointing this out. We agree with this comment.
Therefore, we formalized in the introduction the meaning of background knowledge.
Location in manuscript: Line 52 - 62
Additional clarifications
We added a the Acknowledgments section:
Location in manuscript: Around Line 548 - 549
We changed the Funding section:
Location in manuscript: Around Line 546 – 547
In addition to the above comments, all spelling and grammatical errors pointed out by the reviewers have been corrected. We let the manuscript proofread by a third person for correcting the english.
