arXiv：cs.LG（机器学习，全量分类）· Spyridon Kantarelis, Ioannis Liolitsas, Konstantinos Thomas, Vassilis Lyberatos, Edmund Dervakos, Giorgos Stamou·· 15 小时前AI 评分35
Chordonomicon：66.6 万首歌曲和弦进行数据集发布
CHORDONOMICON: A Dataset of 666,000 Songs and their Chord Progressions
AI 导读
Chordonomicon 发布，包含超 666,000 首歌曲级符号化和弦进行，标注了 verse、chorus、bridge 等结构部分及流派、发行日期。该数据集同时提供可复现基准套件，在严格精确匹配下评估 RNN、GRU、LSTM 三种序列模型，实验显示结构部分标注能持续提升下一和弦预测性能。数据集以开放基准形式发布，含划分方法、基线和评估协议。
正文
Abstract:Chord progressions encapsulate important information about music, pertaining to its structure and conveyed emotions. They serve as the backbone of musical composition, and in many cases, they are the sole information required for a musician to play along and follow the music. Despite their importance, chord progressions as a data domain remain underexplored; existing datasets lack the scale, structural annotation, and metadata diversity required for rigorous evaluation of music understanding models. In this work, we present Chordonomicon, the largest dataset of its kind, containing over 666,000 song-level symbolic chord progressions, annotated with structural parts (verse, chorus, bridge, etc.), genre, and release date, created by scraping various sources of user-generated progressions and associated metadata, showing strong similarity to well-established prior datasets. Beyond the dataset itself, we propose a reproducible benchmark suite for next chord prediction, evaluating three sequence modeling architectures (RNN, GRU, LSTM) across multiple context window sizes and data scales under strict exact-match evaluation. Our experiments reveal that structural part annotations consistently improve prediction performance. Chordonomicon is released as an open benchmark, providing split methodology, baselines, and evaluation protocols to enable fair and reproducible comparison for future work on chord prediction, classification, generation, and beyond.
| Subjects: | Sound (cs.SD); Machine Learning (cs.LG); Multimedia (cs.MM); Audio and Speech Processing (eess.AS) |
| Cite as: | arXiv:2410.22046 [cs.SD] |
| (or arXiv:2410.22046v4 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2410.22046 arXiv-issued DOI via DataCite |
Submission history
From: Spyridon Kantarelis [view email]
[v1]
Tue, 29 Oct 2024 13:53:09 UTC (632 KB)
[v2]
Wed, 27 Nov 2024 09:33:18 UTC (632 KB)
[v3]
Tue, 10 Dec 2024 19:51:42 UTC (1,256 KB)
[v4]
Thu, 1 Oct 2026 08:39:50 UTC (713 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org