Our evaluation methodology adopted the approach for structural segmentation evaluation outlined in the Harmonix set, which employed Structural Features for boundary identification, and 2D-Fourier Magnitude Coefficients (2D-FMC) for segment labeling based on acoustic similarity. CQT features serve as input features for the algorithm. The algorithm is implemented using Music Structure Analysis Framework (MSAF). For evaluation metrics, the F-measure is reported for the following metrics: Hit Rate with 0.5 and 3-second windows for boundary retrieval, Pairwise Frame Clustering and Entropy Scores for segment labeling. The evaluation is implemented using mir_eval.
Evaluation result
Download
By Git
git clone https://www.modelscope.cn/ccmusic-database/song_structure.git
pip install modelscope
By API
from modelscope import snapshot_download
model_dir = snapshot_download('ccmusic-database/song_structure')
Dataset
https://huggingface.co/datasets/ccmusic-database/song_structure
Mirror
https://www.modelscope.cn/models/ccmusic-database/song_structure
Evaluation
https://github.com/monetjoe/ccmusic_eval/tree/msa
Cite
@dataset{zhaorui_liu_2021_5676893,
author = {Monan Zhou, Shenyang Xu, Zhaorui Liu, Zhaowen Wang, Feng Yu, Wei Li and Baoqiang Han},
title = {CCMusic: an Open and Diverse Database for Chinese and General Music Information Retrieval Research},
month = {mar},
year = {2024},
publisher = {HuggingFace},
version = {1.2},
url = {https://huggingface.co/ccmusic-database}
}