Abstract
Current methods for Music Structure Analysis (MSA) focus primarily on audio data. While symbolic music can be synthesized into audio and analyzed using existing MSA techniques, such an approach does not exploit symbolic music's rich explicit representation of pitch, timing, and instrumentation. A key subproblem of MSA is section boundary detection-determining whether a given point in time marks the transition between musical sections. In this paper, we study automatic section boundary detection for symbolic music. First, we introduce a human-annotated MIDI dataset for section boundary detection, consisting of metadata from 6134 MIDI files that we manually curated from the Lakh MIDI dataset. Second, we train a deep learning model to classify the presence of section boundaries within a fixed-length musical window. Our data representation involves a novel encoding scheme based on synthesized overtones to encode arbitrary MIDI instrumentations into 3-channel piano rolls. Our model achieves an F1 score of 0.77, improving over the analogous audio-based supervised learning approach and the unsupervised block-matching segmentation (CBM) audio approach by 0.22 and 0.31, respectively. We release our dataset, code, and models.
Abstract (translated)
目前的音乐结构分析（MSA）方法主要集中在音频数据上。虽然可以通过现有技术将符号音乐合成成音频并进行分析，但这种方法并未充分利用符号音乐在音高、时间以及乐器配置方面的丰富显式表示。MSA的一个关键子问题是节段边界检测——确定某一给定时刻是否标志着音乐部分之间的转换。本文探讨了针对符号音乐的自动节段边界检测。 首先，我们介绍了一个由人工整理的标记有元数据的MIDI数据集，该数据集包含了从Lakh MIDI数据集中手动选取和处理的6134个MIDI文件，用于进行节段边界检测。其次，我们训练了一种深度学习模型，以在固定长度的音乐窗口内对节段边界的出现情况进行分类。我们的数据表示方法采用了基于合成泛音的一种新颖编码方案，将任意MIDI乐器配置编码为三通道钢琴卷帘（piano rolls）。我们的模型实现了0.77的F1得分，相较于同类音频基础监督学习方法和无监督块匹配分割（CBM）音频方法分别提高了0.22和0.31。 我们已发布了数据集、代码及模型。
URL
https://arxiv.org/abs/2509.16566
