sync ms

README.md

@@ -13,16 +13,19 @@ size_categories:
 - n<1K
 viewer: false
 ---
+
 # Dataset Card for Chinese Musical Instruments Timbre Evaluation Database
+The raw dataset encompasses subjective timbre evaluation scores comprising 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 includes 10 reports on spectrogram analysis of 10 instruments.
+
 ## Dataset Description
 - **Homepage:** <https://ccmusic-database.github.io>
 - **Repository:** <https://huggingface.co/datasets/ccmusic-database/CMITE>
 - **Paper:** <https://doi.org/10.5281/zenodo.5676893>
 - **Leaderboard:** <https://ccmusic-database.github.io/team.html>
-- **Point of Contact:** N/A
+- **Point of Contact:** <https://www.modelscope.cn/datasets/ccmusic/instrument_timbre>
 
 ### Dataset Summary
-This database contains 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.
+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
@@ -30,7 +33,30 @@ Musical Instruments Timbre Evaluation
 ### Languages
 Chinese, English
 
+## Usage
+```python
+from datasets import load_dataset
+
+dataset = load_dataset("ccmusic-database/instrument_timbre")
+for item in ds["Chinese"]:
+    print(item)
+
+for item in ds["Non_Chinese"]:
+    print(item)
+```
+
+## Maintenance
+```bash
+GIT_LFS_SKIP_SMUDGE=1 git clone git@hf.co:datasets/ccmusic-database/instrument_timbre
+cd instrument_timbre
+```
+
 ## Dataset Structure
+|         audio(.wav, 22050Hz)         |    mel(.jpg, 22050Hz)     | instrument_name | slim / bright / ... / turbid |
+| :----------------------------------: | :-----------------------: | :-------------: | :--------------------------: |
+| <audio controls src="./data/36.wav"> | <img src="./data/36.jpg"> |     string      |         float(0-10)          |
+|                 ...                  |            ...            |       ...       |             ...              |
+
 ### Data Instances
 .zip(.wav), .csv
 
@@ -107,18 +133,17 @@ SOFTWARE.
 ```
 
 ### Citation Information
-```
+```bibtex
 @dataset{zhaorui_liu_2021_5676893,
-  author       = {Zhaorui Liu, Monan Zhou, Shenyang Xu, Yuan Wang, Zhaowen Wang, Wei Li and Zijin Li},
-  title        = {CCMUSIC DATABASE: A Music Data Sharing Platform for Computational Musicology Research},
-  month        = {nov},
-  year         = {2021},
-  publisher    = {Zenodo},
-  version      = {1.1},
-  doi          = {10.5281/zenodo.5676893},
-  url          = {https://doi.org/10.5281/zenodo.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
+Provide a dataset for musical instruments' timbre evaluation

data/{Non-Chinese.zip → 36.wav}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7dfd960d2293735824bcba1600c362a454bac77d4090d7e47db554c22f6c8a33
-size 70292202
+oid sha256:6fb2599805b354e5bf43d17e0fb48a839f9f720ed9d25e0609a9ecf20e64279d
+size 719756

data/Chinese.csv

@@ -1,38 +0,0 @@
-instrument_id,instrument_name,slim,bright,dim,sharp,thick,thin,solid,clear,dry,plump,rough,pure,hoarse,harmonious,soft,turbid
-1,gao_hu,6.470588235,6.676470588,3.411764706,7.176470588,2.764705882,6.588235294,3.176470588,6.088235294,6.823529412,3.205882353,5.617647059,4.676470588,5.617647059,4.647058824,3.441176471,4.323529412
-2,er_hu,5.970588235,6.176470588,3.823529412,5.970588235,3.441176471,5.588235294,4.117647059,5.441176471,5.264705882,4.264705882,4.382352941,4.882352941,4.911764706,6.176470588,5.205882353,4.323529412
-3,zhong_hu,4.411764706,4.941176471,4.852941176,4.882352941,5.058823529,4.588235294,5.088235294,4.882352941,5.294117647,5.205882353,5.117647059,5.029411765,4.941176471,5.852941176,4.882352941,4.441176471
-4,ge_hu,2.647058824,2.970588235,6.764705882,2.617647059,7.588235294,2.588235294,7.558823529,2.705882353,3.235294118,7.352941176,4.735294118,5.647058824,4.117647059,7.088235294,7.176470588,6.147058824
-5,di_yin_ge_hu,1.676470588,1.470588235,8.088235294,1.794117647,8.764705882,1.558823529,8.647058824,1.117647059,4.205882353,7.294117647,6.941176471,3,6.794117647,3.617647059,3.323529412,8
-6,jing_hu,7.588235294,7.676470588,2.205882353,8.058823529,1.970588235,7.205882353,2.705882353,7.882352941,5.823529412,3.852941176,3.323529412,6.235294118,3.823529412,5,3.676470588,2.441176471
-7,ban_hu,7.441176471,7.529411765,2.264705882,8.323529412,2.088235294,7.205882353,2.411764706,7.558823529,6.529411765,2.941176471,5.735294118,4.705882353,6.264705882,3.941176471,2.823529412,3.294117647
-8,bang_di,8.323529412,8.588235294,1.5,8.411764706,1.411764706,7.970588235,1.529411765,8.735294118,5.147058824,3.117647059,3.029411765,6.705882353,3.382352941,5.647058824,2.970588235,1.911764706
-9,qu_di,6.264705882,6.878787879,2.970588235,6.294117647,3.176470588,6.029411765,3.382352941,6.911764706,4.294117647,4.147058824,2.764705882,6.617647059,3.441176471,7.058823529,5.294117647,2.558823529
-10,xin_di,4.764705882,4.647058824,5.272727273,4.176470588,4.852941176,4.735294118,4.558823529,4.705882353,4,4.911764706,4.558823529,5.264705882,5.029411765,5.882352941,6.205882353,5.235294118
-11,da_di,5.058823529,5.735294118,4.205882353,4.588235294,4.411764706,4.941176471,4.441176471,5.411764706,3.382352941,5.235294118,3.823529412,6.029411765,3.852941176,6.588235294,6.176470588,4.676470588
-12,gao_yin_sheng,5.617647059,6.117647059,3.96969697,5.882352941,4.235294118,5.941176471,3.970588235,5.147058824,4.970588235,4.5,4.470588235,5.529411765,4.441176471,5.176470588,4.764705882,4.352941176
-13,zhong_yin_sheng,3.794117647,3.617647059,6.176470588,3.852941176,6.323529412,4.117647059,6.029411765,3.205882353,4.058823529,5.5,4.441176471,5.03030303,3.941176471,5.382352941,5.735294118,5.647058824
-14,di_yin_sheng,2.882352941,3.029411765,6.588235294,2.970588235,6.764705882,3.529411765,6.676470588,2.705882353,3.764705882,6.176470588,5,4.088235294,4.558823529,5.470588235,4.705882353,6.441176471
-15,gao_yin_suo_na,7.176470588,7.588235294,2.794117647,7.735294118,2.441176471,7.264705882,2.941176471,7.352941176,6.441176471,4.333333333,4.705882353,5.242424242,5.323529412,3.941176471,2.794117647,3.235294118
-16,zhong_yin_suo_na,4.352941176,5.911764706,4.352941176,6.117647059,3.970588235,6.205882353,4.588235294,5.029411765,7.029411765,4.558823529,5.588235294,4.705882353,5.882352941,3.617647059,2.764705882,4.117647059
-17,ci_zhong_yin_suo_na,3.794117647,4.852941176,5.558823529,5.352941176,4.735294118,5.852941176,4.941176471,3.705882353,7.529411765,4.411764706,6.294117647,4.088235294,6.5,2.676470588,2.470588235,5.058823529
-18,di_yin_suo_na,4.845371972,5.543691675,6.411764706,4.411764706,6.147058824,4.294117647,6.529411765,2.676470588,6.941176471,5.205882353,6.676470588,3.794117647,6.352941176,3.029411765,2.205882353,6.205882353
-19,gao_yin_guan,5.941176471,6,4.470588235,7.235294118,3.5,6.352941176,3.470588235,5.323529412,7.676470588,3.529411765,6.911764706,2.794117647,7.382352941,2.147058824,2.147058824,5.323529412
-20,zhong_yin_guan,4.794117647,5.382352941,4.852941176,5.029411765,5.088235294,4.764705882,5,4.5,4.5,5.382352941,5.176470588,4.852941176,4.970588235,5.058823529,5.264705882,4.823529412
-21,di_yin_guan,3.088235294,3.176470588,6.176470588,3.764705882,7.029411765,3.529411765,7.235294118,2.735294118,4.176470588,6.617647059,5.882352941,4.411764706,6,5.117647059,4.441176471,6.794117647
-22,bei_di_yin_guan,2.176470588,2.235294118,7.147058824,3.235294118,7.911764706,2.941176471,7.941176471,2.058823529,4.352941176,6.705882353,6.911764706,3.029411765,6.382352941,4.205882353,3.470588235,7.617647059
-23,ba_wu,5.181818182,5.558823529,3.882352941,4.264705882,4.470588235,4.411764706,4.735294118,6.088235294,2.735294118,5.794117647,3.058823529,6.794117647,3.470588235,6.970588235,5.735294118,3.588235294
-24,xun,4.647058824,3.323529412,6.529411765,3.205882353,5.147058824,4.911764706,4.323529412,3.382352941,4.676470588,4.5,5.794117647,4.176470588,6.823529412,4.470588235,6.264705882,6.5
-25,xiao,4.882352941,4.382352941,5.676470588,3.5,5.205882353,4.529411765,4.705882353,4.058823529,3.794117647,5.352941176,3.941176471,5.911764706,4.411764706,6.794117647,7.147058824,5.235294118
-26,liu_qin,6.882352941,7.323529412,3,6.352941176,3.382352941,6.029411765,3.088235294,7.5,4,5.235294118,3.441176471,5.941176471,3.058823529,6.363636364,4.882352941,3.058823529
-27,xiao_ruan,6.882352941,7.382352941,2.911764706,6.264705882,3.117647059,6.176470588,2.970588235,7.647058824,3.647058824,5.088235294,3.264705882,6.147058824,3.294117647,6.588235294,5,2.794117647
-28,pi_pa,7,6.941176471,3.323529412,5.794117647,3.264705882,6.352941176,2.705882353,7.205882353,4.323529412,4.235294118,4.117647059,5.235294118,4.176470588,5.705882353,4.411764706,3.352941176
-29,yang_qin,5.794117647,6.5,3.176470588,5.058823529,4.235294118,4.647058824,4.147058824,6.529411765,2.852941176,5.676470588,3.617647059,5.323529412,3.176470588,6.647058824,5.205882353,5.147058824
-30,zhong_ruan,4.470588235,5.264705882,4.941176471,3.794117647,6.147058824,3.588235294,6.088235294,5.294117647,2.882352941,6.647058824,4.147058824,5.676470588,3.176470588,6.529411765,5.617647059,5.352941176
-31,da_ruan,3.088235294,4.088235294,5.823529412,2.588235294,7.382352941,2.676470588,7.5,3.705882353,2.029411765,7.705882353,3.411764706,6.764705882,2.411764706,7.529411765,6.911764706,5.529411765
-32,gu_zheng,5.764705882,6.441176471,3.529411765,5.176470588,4.176470588,5.382352941,4.205882353,6.558823529,3.617647059,5.852941176,3.529411765,5.941176471,3.176470588,6.294117647,5.212121212,3.764705882
-33,gu_qin,3.764705882,4,5.735294118,3.882352941,6.205882353,4.147058824,6.088235294,4.088235294,4.235294118,6.352941176,5.382352941,4.5,4.852941176,4.735294118,5.121212121,5.764705882
-34,kong_hou,5.088235294,5.441176471,4.294117647,3.470588235,5.941176471,3.382352941,5.941176471,5.235294118,2.617647059,7.088235294,3.205882353,6.941176471,2.205882353,7.647058824,7.294117647,4.5
-35,san_xian,5.352941176,4.823529412,4.676470588,4.676470588,3.705882353,6.264705882,3.647058824,5.411764706,6.117647059,3.676470588,6.470588235,3.352941176,5.676470588,4.382352941,3.5,5
-36,yun_luo,6.441176471,6.735294118,3.441176471,6.382352941,2.764705882,7.205882353,2.588235294,7.382352941,5.176470588,2.794117647,4.764705882,5.882352941,4.235294118,4.029411765,4.029411765,4.264705882
-37,bian_zhong,5.205882353,4.882352941,5,4.794117647,4.705882353,5.5,4.205882353,5.852941176,4.676470588,3.882352941,5,5.794117647,4.588235294,4.647058824,5.088235294,5.941176471

data/Non-Chinese.csv

@@ -1,25 +0,0 @@
-instrument_id,instrument_name,slim,bright,dim,sharp,thick,thin,solid,clear,dry,plump,rough,pure,hoarse,harmonious,soft,turbid
-1,violin,5.2,5.3,3.4,4.1,4.1,3.7,5,4.5,3.9,6.4,3.5,6,4.1,4.5,6.4,6
-2,viola,3.5,4.4,4.4,3.4,6.2,2.9,6.3,3.6,3.3,7.1,4.2,5.6,4.5,5.2,6.3,6.8
-3,cello,2.7,3.1,5.5,2.3,7.6,1.9,7.6,2.3,2.8,7.8,4.7,6.3,4.5,6,6.8,7.2
-4,double_bass,1.6,1,8.5,1.3,8.8,1.8,8.7,1.1,3.2,6.6,7.6,3.9,6.4,8.1,3,3.8
-5,piccolo,7.2,8,2,7.9,2.5,7.1,3,7.5,3.7,3.5,3.1,5.5,4.1,3.6,5.9,5.1
-6,flute,6.6,6.5,2.9,6.9,3.4,6,3,6.9,3.6,4.7,2.7,5.8,2.8,3.3,6.7,6.6
-7,oboe,5.9,6.5,2.9,6.1,3.1,6.4,3.7,5.7,4.1,5.3,3.2,6,3.6,3.5,6.4,6.4
-8,clarinet,4.2,4.8,4.5,4.3,5.9,4.5,6.2,3.9,3.6,6.5,3.7,6.6,3.5,5,7.2,7.2
-9,bassoon,2.1,2.3,7.1,2.3,7.7,4.5,8.1,1.9,5.7,6.4,7.4,3,7,7.6,3.1,3.9
-10,saxophone,3.2,4.3,6.1,3.3,7.3,4.3,6.9,3.1,4.6,5.6,6.6,5,6.3,7.2,3.6,5.1
-11,trumpet,5.5,7.1,2.4,7.2,3,6,3.8,6.1,5.7,4.2,6.3,3.4,5.8,3.9,3.6,2.1
-12,trombone,3.5,4.8,5,4.7,5.6,3.4,6.7,3.8,4.8,6.2,6.5,4.2,4.9,5.7,5,4.2
-13,horn,2.4,2.9,6.5,3.6,7.4,3,7.6,2.4,6.1,6,6.6,4.1,5.9,7.7,4.3,4.3
-14,tuba,1.2,1,8.7,1.5,8.6,1.6,8.6,1,5.7,7.4,7.6,3.5,7.1,8.6,5,5.2
-15,harp,5.3,5.6,3.8,4,4.1,4.2,5,6,2.9,6.1,2.7,7.5,2.9,3.9,7.2,6.9
-16,tubular_bells,5.5,6.7,2.4,6.1,2.7,7.4,2.5,7.6,6.3,3.2,6.1,3,4.3,5.9,2.5,2.1
-17,bells,8.8,8.7,1,8.5,1,8.7,1,9,5.7,2.5,1.7,8.6,2.3,1.3,7.2,3.4
-18,xylophone,6,6.5,2.4,6.8,1.8,8,2,8,5.2,3,2.9,7.9,2.1,1.9,5.4,4.1
-19,vibraphone,5.5,6.9,2,5.9,2.7,6.3,3.5,7.8,3.2,5.5,2.3,7.1,2.1,3.9,7.2,6
-20,marimba,3.8,4.6,4.8,4.6,5,4.6,5.7,5.7,4.7,5,3.4,5.8,3.4,4.4,6.3,6
-21,piano,3,4.5,4.8,3.7,5.9,3.5,6.1,5.1,3.6,6.8,3.9,6.8,3.3,4.3,7.6,6.1
-22,clavichord,3.8,5.9,3.4,6.7,4.1,6.8,3.9,5.3,5.2,4.4,6.6,2.8,7.9,6.9,3.5,2.1
-23,accordion,4.7,6.5,2.6,5.8,4.6,5.3,4.3,5.1,3.2,6.6,4.9,5.1,3.8,4.8,7.2,4.5
-24,organ,3.1,6.3,3.5,5.7,6.6,4.3,5.9,2.8,3.6,7.6,4.9,4.1,5.4,7.5,5.6,4.5

instrument_timbre.py

@@ -6,41 +6,99 @@ from datasets.tasks import AudioClassification
 
 _NAMES = [
     # Chinese
-    'gao_hu', 'er_hu', 'zhong_hu', 'ge_hu', 'di_yin_ge_hu', 'jing_hu', 'ban_hu', 'bang_di', 'qu_di', 'xin_di', 'da_di',
-    'gao_yin_sheng', 'zhong_yin_sheng', 'di_yin_sheng', 'gao_yin_suo_na', 'zhong_yin_suo_na', 'ci_zhong_yin_suo_na',
-    'di_yin_suo_na', 'gao_yin_guan', 'zhong_yin_guan', 'di_yin_guan', 'bei_di_yin_guan', 'ba_wu', 'xun', 'xiao', 'liu_qin',
-    'xiao_ruan', 'pi_pa', 'yang_qin', 'zhong_ruan', 'da_ruan', 'gu_zheng', 'gu_qin', 'kong_hou', 'san_xian', 'yun_luo',
-    'bian_zhong',
+    "gao_hu",
+    "er_hu",
+    "zhong_hu",
+    "ge_hu",
+    "di_yin_ge_hu",
+    "jing_hu",
+    "ban_hu",
+    "bang_di",
+    "qu_di",
+    "xin_di",
+    "da_di",
+    "gao_yin_sheng",
+    "zhong_yin_sheng",
+    "di_yin_sheng",
+    "gao_yin_suo_na",
+    "zhong_yin_suo_na",
+    "ci_zhong_yin_suo_na",
+    "di_yin_suo_na",
+    "gao_yin_guan",
+    "zhong_yin_guan",
+    "di_yin_guan",
+    "bei_di_yin_guan",
+    "ba_wu",
+    "xun",
+    "xiao",
+    "liu_qin",
+    "xiao_ruan",
+    "pi_pa",
+    "yang_qin",
+    "zhong_ruan",
+    "da_ruan",
+    "gu_zheng",
+    "gu_qin",
+    "kong_hou",
+    "san_xian",
+    "yun_luo",
+    "bian_zhong",
     # Non-Chinese
-    'violin', 'viola', 'cello', 'double_bass', 'piccolo', 'flute', 'oboe', 'clarinet', 'bassoon', 'saxophone', 'trumpet',
-    'trombone', 'horn', 'tuba', 'harp', 'tubular_bells', 'bells', 'xylophone', 'vibraphone', 'marimba', 'piano',
-    'clavichord', 'accordion', 'organ'
+    "violin",
+    "viola",
+    "cello",
+    "double_bass",
+    "piccolo",
+    "flute",
+    "oboe",
+    "clarinet",
+    "bassoon",
+    "saxophone",
+    "trumpet",
+    "trombone",
+    "horn",
+    "tuba",
+    "harp",
+    "tubular_bells",
+    "bells",
+    "xylophone",
+    "vibraphone",
+    "marimba",
+    "piano",
+    "clavichord",
+    "accordion",
+    "organ",
 ]
 
-_HOMEPAGE = f"https://huggingface.co/datasets/ccmusic-database/{os.path.basename(__file__).split('.')[0]}"
+_DBNAME = os.path.basename(__file__).split(".")[0]
+
+_DOMAIN = f"https://www.modelscope.cn/api/v1/datasets/ccmusic/{_DBNAME}/repo?Revision=master&FilePath=data"
+
+_HOMEPAGE = f"https://www.modelscope.cn/datasets/ccmusic/{_DBNAME}"
 
 _CITATION = """\
 @dataset{zhaorui_liu_2021_5676893,
-  author       = {Zhaorui Liu, Monan Zhou, Shenyang Xu, Yuan Wang, Zhaowen Wang, Wei Li and Zijin Li},
-  title        = {CCMUSIC DATABASE: A Music Data Sharing Platform for Computational Musicology Research},
-  month        = {nov},
-  year         = {2021},
-  publisher    = {Zenodo},
-  version      = {1.1},
-  doi          = {10.5281/zenodo.5676893},
-  url          = {https://doi.org/10.5281/zenodo.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}
 }
 """
 
 _DESCRIPTION = """\
-This database contains 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. Furthermore, 10 reports on spectrum analysis of 10 instruments are also included.
+The raw dataset encompasses subjective timbre evaluation scores comprising 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 includes 10 reports on spectrogram analysis of 10 instruments.
+
+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 44,100 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.
 """
 
 _URLS = {
-    'cn_zip': f"{_HOMEPAGE}/resolve/main/data/Chinese.zip",
-    'cn_csv': f"{_HOMEPAGE}/resolve/main/data/Chinese.csv",
-    'zip': f"{_HOMEPAGE}/resolve/main/data/Non-Chinese.zip",
-    'csv': f"{_HOMEPAGE}/resolve/main/data/Non-Chinese.csv"
+    "audio": f"{_DOMAIN}/audio.zip",
+    "mel": f"{_DOMAIN}/mel.zip",
+    "csv": f"{_DOMAIN}/Chinese.csv",
+    "label": f"{_DOMAIN}/Non-Chinese.csv",
 }
 
 
@@ -48,26 +106,29 @@ class instrument_timbre(datasets.GeneratorBasedBuilder):
     def _info(self):
         return datasets.DatasetInfo(
             description=_DESCRIPTION,
-            features=datasets.Features({
-                "audio": datasets.Audio(sampling_rate=44_100),
-                "instrument_name": datasets.features.ClassLabel(names=_NAMES),
-                "slim": datasets.Value("float64"),
-                "bright": datasets.Value("float64"),
-                "dim": datasets.Value("float64"),
-                "sharp": datasets.Value("float64"),
-                "thick": datasets.Value("float64"),
-                "thin": datasets.Value("float64"),
-                "solid": datasets.Value("float64"),
-                "clear": datasets.Value("float64"),
-                "dry": datasets.Value("float64"),
-                "plump": datasets.Value("float64"),
-                "rough": datasets.Value("float64"),
-                "pure": datasets.Value("float64"),
-                "hoarse": datasets.Value("float64"),
-                "harmonious": datasets.Value("float64"),
-                "soft": datasets.Value("float64"),
-                "turbid": datasets.Value("float64")
-            }),
+            features=datasets.Features(
+                {
+                    "audio": datasets.Audio(sampling_rate=22050),
+                    "mel": datasets.Image(),
+                    "instrument_name": datasets.features.ClassLabel(names=_NAMES),
+                    "slim": datasets.Value("float64"),
+                    "bright": datasets.Value("float64"),
+                    "dim": datasets.Value("float64"),
+                    "sharp": datasets.Value("float64"),
+                    "thick": datasets.Value("float64"),
+                    "thin": datasets.Value("float64"),
+                    "solid": datasets.Value("float64"),
+                    "clear": datasets.Value("float64"),
+                    "dry": datasets.Value("float64"),
+                    "plump": datasets.Value("float64"),
+                    "rough": datasets.Value("float64"),
+                    "pure": datasets.Value("float64"),
+                    "hoarse": datasets.Value("float64"),
+                    "harmonious": datasets.Value("float64"),
+                    "soft": datasets.Value("float64"),
+                    "turbid": datasets.Value("float64"),
+                }
+            ),
             supervised_keys=("audio", "instrument_name"),
             homepage=_HOMEPAGE,
             license="mit",
@@ -76,73 +137,76 @@ class instrument_timbre(datasets.GeneratorBasedBuilder):
                 AudioClassification(
                     task="audio-classification",
                     audio_column="audio",
-                    label_column="instrument_name"
+                    label_column="instrument_name",
                 )
-            ]
+            ],
         )
 
-    def _custom_sort(self, item):
-        return int(os.path.basename(item).split('.wa')[0])
-
     def _split_generators(self, dl_manager):
-        cn_data_files = dl_manager.download_and_extract(_URLS['cn_zip'])
-        data_files = dl_manager.download_and_extract(_URLS['zip'])
-
-        cn_files = dl_manager.iter_files([cn_data_files])
-        files = dl_manager.iter_files([data_files])
-
-        cn_ins_eval = dl_manager.download(_URLS['cn_csv'])
-        ins_eval = dl_manager.download(_URLS['csv'])
-
-        cn_labels = pd.read_csv(cn_ins_eval, index_col='instrument_id')
-        labels = pd.read_csv(ins_eval, index_col='instrument_id')
-
-        cn_dataset, dataset = [], []
-        for cpath in cn_files:
-            if os.path.basename(cpath).endswith(".wav"):
-                cn_dataset.append(cpath)
-
-        for epath in files:
-            if os.path.basename(epath).endswith(".wav"):
-                dataset.append(epath)
+        audio_files = dl_manager.download_and_extract(_URLS["audio"])
+        mel_files = dl_manager.download_and_extract(_URLS["mel"])
+        cn_ins_eval = dl_manager.download(_URLS["csv"])
+        ins_eval = dl_manager.download(_URLS["label"])
+
+        cn_labels = pd.read_csv(cn_ins_eval, index_col="instrument_id")
+        labels = pd.read_csv(ins_eval, index_col="instrument_id")
+        cn_dataset, dataset = {}, {}
+
+        for path in dl_manager.iter_files([audio_files]):
+            fname = os.path.basename(path)
+            ins_id = int(fname.split(".wa")[0]) - 1
+            if fname.endswith(".wav"):
+                if os.path.basename(os.path.dirname(path)) == "Chinese":
+                    cn_dataset[ins_id] = {"audio": path}
+                else:
+                    dataset[ins_id] = {"audio": path}
+
+        for path in dl_manager.iter_files([mel_files]):
+            fname = os.path.basename(path)
+            ins_id = int(fname.split(".jp")[0]) - 1
+            if fname.endswith(".jpg"):
+                if os.path.basename(os.path.dirname(path)) == "Chinese":
+                    cn_dataset[ins_id]["mel"] = path
+                else:
+                    dataset[ins_id]["mel"] = path
 
         return [
             datasets.SplitGenerator(
                 name="Chinese",
                 gen_kwargs={
-                    "files": sorted(cn_dataset, key=self._custom_sort),
-                    "labels": cn_labels
-                }
+                    "files": [cn_dataset[k] for k in sorted(cn_dataset)],
+                    "labels": cn_labels,
+                },
             ),
             datasets.SplitGenerator(
                 name="Non_Chinese",
                 gen_kwargs={
-                    "files": sorted(dataset, key=self._custom_sort),
-                    "labels": labels
-                }
-            )
+                    "files": [dataset[k] for k in sorted(dataset)],
+                    "labels": labels,
+                },
+            ),
         ]
 
-    def _generate_examples(self, files, labels):
-        for path in files:
-            i = int(os.path.basename(path).split('.wa')[0]) - 1
+    def _generate_examples(self, files, labels: pd.DataFrame):
+        for i, path in enumerate(files):
             yield i, {
-                "audio": path,
-                "instrument_name": labels.iloc[i]['instrument_name'],
-                "slim": labels.iloc[i]['slim'],
-                "bright": labels.iloc[i]['bright'],
-                "dim": labels.iloc[i]['dim'],
-                "sharp": labels.iloc[i]['sharp'],
-                "thick": labels.iloc[i]['thick'],
-                "thin": labels.iloc[i]['thin'],
-                "solid": labels.iloc[i]['solid'],
-                "clear": labels.iloc[i]['clear'],
-                "dry": labels.iloc[i]['dry'],
-                "plump": labels.iloc[i]['plump'],
-                "rough": labels.iloc[i]['rough'],
-                "pure": labels.iloc[i]['pure'],
-                "hoarse": labels.iloc[i]['hoarse'],
-                "harmonious": labels.iloc[i]['harmonious'],
-                "soft": labels.iloc[i]['soft'],
-                "turbid": labels.iloc[i]['turbid']
+                "audio": path["audio"],
+                "mel": path["mel"],
+                "instrument_name": labels.iloc[i]["instrument_name"],
+                "slim": labels.iloc[i]["slim"],
+                "bright": labels.iloc[i]["bright"],
+                "dim": labels.iloc[i]["dim"],
+                "sharp": labels.iloc[i]["sharp"],
+                "thick": labels.iloc[i]["thick"],
+                "thin": labels.iloc[i]["thin"],
+                "solid": labels.iloc[i]["solid"],
+                "clear": labels.iloc[i]["clear"],
+                "dry": labels.iloc[i]["dry"],
+                "plump": labels.iloc[i]["plump"],
+                "rough": labels.iloc[i]["rough"],
+                "pure": labels.iloc[i]["pure"],
+                "hoarse": labels.iloc[i]["hoarse"],
+                "harmonious": labels.iloc[i]["harmonious"],
+                "soft": labels.iloc[i]["soft"],
+                "turbid": labels.iloc[i]["turbid"],
             }