Emotion4MIDI
收藏资源简介:
Emotion4MIDI是一个基于歌词的大规模情感标记符号音乐数据集,由系统与计算机工程技术科学研究所创建。该数据集包含12509首MIDI歌曲,通过分析歌词来标记情感,覆盖了广泛的精细情感范围。创建过程中,研究者首先在GoEmotions数据集上训练情感分类模型,然后将这些模型应用于两个大型MIDI数据集的歌词上。Emotion4MIDI数据集旨在探索音乐与情感之间的联系,特别是开发基于特定情感生成音乐的模型,为情感算法作曲领域的研究提供了宝贵的资源。
Emotion4MIDI is a large-scale lyric-based sentiment-labeled symbolic music dataset developed by the Institute of Systems, Computer Engineering, and Technical Sciences. This dataset comprises 12,509 MIDI songs, with sentiment labels assigned through lyric analysis, covering a broad spectrum of fine-grained emotional categories. During its construction, researchers first trained sentiment classification models on the GoEmotions dataset, then deployed these models to analyze the lyrics of two large-scale MIDI datasets. The Emotion4MIDI dataset aims to investigate the association between music and emotion, particularly to facilitate the development of models that generate music conditioned on specific emotional states, thus providing a valuable resource for research in the field of sentiment-aware algorithmic composition.

- 1Emotion4MIDI: a Lyrics-based Emotion-Labeled Symbolic Music Dataset系统与计算机工程技术科学研究所 · 2023年



