five

anusfoil/ycuppe-midi

收藏
Hugging Face2026-03-27 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/anusfoil/ycuppe-midi
下载链接
链接失效反馈
官方服务:
资源简介:
--- license: cc-by-4.0 task_categories: - audio-classification tags: - piano - midi - performance-assessment - music size_categories: - 1K<n<10K --- # YCU-PPE-III: Piano Performance MIDI Dataset MIDI transcriptions of the YCU-PPE-III piano performance dataset (Wang et al.), used for **unreferenced Performance MOS (PMOS) prediction** in [EVPMR](https://github.com/anusfoil/eval-piano-midi-repr). ## Overview - **2,627 MIDI files** transcribed from WAV recordings via [transkun](https://github.com/RongTongZi/transkun) - **13 songs** performed by student pianists - **2,511 performances** with ratings from 3 expert judges (0-100 scale each) - **Labels**: normalized mean score to [0, 1] - **Splits**: 1,757 train / 377 val / 377 test (stratified by song) ## Metadata `metadata/metadata.csv` columns: midi_path, name, song_id, piece_cn, piece_en, judge_1, judge_2, judge_3, mean_score, label, split ## Task Regression: given a piano performance MIDI (without reference score), predict a quality rating. Metrics: MSE, MAE, Pearson r, Spearman rho. ## Directory Structure ``` data/{song_id}/{name}.mid # 2,627 MIDI files metadata/splits.csv # 2,511 entries: midi_path, label, split metadata/metadata.csv # Full metadata with judge scores + piece names ```
提供机构:
anusfoil
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作