There is an increasing interest in the automatic digitization of medieval music documents. Despite efforts in this field, the detection of the different layers of information on these documents still poses difficulties. The use of Deep Neural Networks techniques has reported outstanding results in many areas related to computer vision. Consequently, in this paper, we study the so-called Convolutional Neural Networks (CNN) for performing the automatic document processing of music score images. This process is focused on layering the image into its constituent parts (namely, background, staff lines, music notes, and text) by training a classifier with examples of these parts. A comprehensive experimentation in terms of the configuration of the networks was carried out, which illustrates interesting results as regards to both the efficiency and effectiveness of these models. In addition, a cross-manuscript adaptation experiment was presented in which the networks are evaluated on a different manuscript from the one they were trained. The results suggest that the CNN is capable of adapting its knowledge, and so starting from a pre-trained CNN reduces (or eliminates) the need for new labeled data.

Optical music recognition (OMR) is the field that studies how to automatically read music notation from score images. One of the relevant steps within the OMR workflow is the staff-region retrieval. This process is a key step because any undetected staff will not be processed by the subsequent steps. This task has previously been addressed as a supervised learning problem in the literature; however, ground-truth data are not always available, so each new manuscript requires a preliminary manual annotation. This situation is one of the main bottlenecks in OMR, because of the countless number of existing manuscripts , and the associated manual labeling cost. With the aim of mitigating this issue, we propose the application of a domain adaptation technique, the so-called Domain-Adversarial Neural Network (DANN), based on a combination of a gradient reversal layer and a domain classifier in the inference neural architecture. The results from our experiments support the benefits of our proposed solution, obtaining improvements of approximately 29% in the F-score.

Optical music recognition (OMR ) describes the process of automatically transcribing music notation from a digital image. Although similar to optical character recognition (OCR ), the process and procedures of OMR diverge due to the fundamental differences between text and music notation, such as the two-dimensional nature of the notation system and the overlay of music symbols on top of staff lines. The OMR process can be described as a sequence of steps, with techniques adapted from disciplines including image processing, machine learning, grammars, and notation encoding. The sequence and specific techniques used can differ depending on the condition of the image, the type of notation, and the desired output. Several commercial and open-source OMR software systems have been available since the mid-1990s. Most of them are designed to be used by individuals and recognize common (post-18th-century) Western music notation, though there have been some efforts to recognize other types of music notation such as for the lute and for earlier Western music. Even though traditional applications of OMR have focused on small-scale recognition tasks, typically as an automated method of musical entry for score editing, new applications of large-scale OMR are under development, where automated recognition is the central technology for building full-music search systems, similar to the large-scale full-text recognition efforts.

The automatic analysis of scores has been a research topic of interest for the last few decades and still is since music databases that include musical scores are currently being created to make musical content available to the public, including scores of ancient music. For the correct analysis of music elements and their interpretation, the identification of staff lines is of key importance. In this paper, a scheme to post-process the output of a previous musical object identification system is described. This system allows the reconstruction by means of detection, tracking and interpolation of the staff lines of ancient scores from the digital Salzinnes Database. The scheme developed shows a remarkable performance on the specific task it was created for.

This study evaluates the baseline capabilities of Large Language Models (LLMs) like ChatGPT, Claude, and Gemini to learn concepts in music theory through in-context learning and chain-of-thought prompting. Using carefully designed prompts (in-context learning) and step-by-step worked examples (chain-of-thought prompting), we explore how LLMs can be taught increasingly complex material and how pedagogical strategies for human learners translate to educating machines. Performance is evaluated using questions from an official Canadian Royal Conservatory of Music (RCM) Level 6 examination, which covers a comprehensive range of topics, including interval and chord identification, key detection, cadence classification, and metrical analysis. Additionally, we evaluate the suitability of various music encoding formats for these tasks (ABC, Humdrum, MEI, MusicXML). All experiments were run both with and without contextual prompts. Results indicate that without context, ChatGPT with MEI performs the best at 52%, while with context, Claude with MEI performs the best at 75%. Future work will further refine prompts and expand to cover more advanced music theory concepts. This research contributes to the broader understanding of teaching LLMs and has applications for educators, students, and developers of AI music tools alike.

Recent progress in text-based Large Language Models (LLMs) and their extended ability to process multi-modal sensory data have led us to explore their applicability in addressing music information retrieval (MIR) challenges. In this paper, we use a systematic prompt engineering approach for LLMs to solve MIR problems. We convert the music data to symbolic inputs and evaluate LLMs' ability in detecting annotation errors in three key MIR tasks: beat tracking, chord extraction, and key estimation. A concept augmentation method is proposed to evaluate LLMs' music reasoning consistency with the provided music concepts in the prompts. Our experiments tested the MIR capabilities of Generative Pre-trained Transformers (GPT). Results show that GPT has an error detection accuracy of 65.20%, 64.80%, and 59.72% in beat tracking, chord extraction, and key estimation tasks, respectively, all exceeding the random baseline. Moreover, we observe a positive correlation between GPT's error finding accuracy and the amount of concept information provided. The current findings based on symbolic music input provide a solid ground for future LLM-based MIR research.