Dataset Card for Chinese Musical Instruments Timbre Evaluation Database
The original dataset is sourced from the National Musical Instruments Timbre Evaluation Dataset, which includes subjective timbre evaluation scores using 16 terms such as bright, dark, raspy, etc., evaluated across 37 Chinese instruments and 24 Western instruments by 14 participants with musical backgrounds in a subjective evaluation experiment. Additionally, it contains 10 spectrogram analysis reports for 10 instruments.
Based on the aforementioned original dataset, after data processing, we have constructed the default subset of the current integrated version of the dataset, dividing the Chinese section and the Western section into two splits. Each split consists of multiple data entries, with each entry structured across 18 columns. The Chinese split includes 37 entries, while the Western split comprises 24 entries. The first column of each data entry presents the instrument recordings in .wav format, sampled at a rate of 44,100 Hz. The second column provides the Chinese pinyin or English name of the instrument. The following 16 columns correspond to the 10-point scores of the 16 terms. This dataset is suitable for conducting timbre analysis of musical instruments and can also be utilized for various single or multiple regression tasks related to term scoring. The data structure of the default subset can be viewed in the viewer. As the current dataset has been cited and used in two published articles, the construction of eval subset has been omitted.
Viewer
https://www.modelscope.cn/datasets/ccmusic-database/instrument_timbre/dataPeview
Dataset Structure
audio | mel | instrument_name | slim / bright / ... / turbid (16 colums) |
---|---|---|---|
.wav, 44100Hz | .jpg, 44100Hz | string | float(0-10) |
... | ... | ... | ... |
Data Instances
.zip(.wav), .csv
Data Fields
Chinese traditional instruments / Western instruments
Data Splits
Chinese, Western
Dataset Summary
During the integration, we have crafted the Chinese part and the Non-Chinese part into two splits. Each split is composed of multiple data entries, with each entry structured across 18 columns. The Chinese split encompasses 37 entries, while the Non-Chinese split includes 24 entries. The premier column of each data entry presents the instrument recordings in the .wav format, sampled at a rate of 22,050 Hz. The second column provides the Chinese pinyin or English name of the instrument. The subsequent 16 columns correspond to the 10-point score of the 16 terms. This dataset is suitable for conducting timber analysis of musical instruments and can also be utilized for various single or multiple regression tasks related to term scoring.
Supported Tasks and Leaderboards
Musical Instruments Timbre Evaluation
Languages
Chinese, English
Usage
from datasets import load_dataset
ds = load_dataset("ccmusic-database/instrument_timbre")
for item in ds["Chinese"]:
print(item)
for item in ds["Western"]:
print(item)
Maintenance
git clone git@hf.co:datasets/ccmusic-database/instrument_timbre
cd instrument_timbre
Dataset Creation
Curation Rationale
Lack of a dataset for musical instruments timbre evaluation
Source Data
Initial Data Collection and Normalization
Zhaorui Liu, Monan Zhou
Who are the source language producers?
Students from CCMUSIC
Annotations
Annotation process
Subjective timbre evaluation scores of 16 subjective timbre evaluation terms (such as bright, dark, raspy) on 37 Chinese national and 24 Non-Chinese terms given by 14 participants in a subjective evaluation experiment
Who are the annotators?
Students from CCMUSIC
Personal and Sensitive Information
None
Considerations for Using the Data
Social Impact of Dataset
Promoting the development of AI in the music industry
Discussion of Biases
Only for traditional instruments
Other Known Limitations
Limited data
Additional Information
Dataset Curators
Zijin Li
Evaluation
For Chinese instruments
For Non-Chinese instruments
Citation Information
@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}
}
Contributions
Provide a dataset for musical instruments' timbre evaluation
- Downloads last month
- 115