music_genre.py

import os
import random
import datasets
from datasets.tasks import ImageClassification

_NAMES_1 = {
    1: "Classic",
    2: "Non_classic"
}

_NAMES_2 = {
    3: "Symphony",
    4: "Opera",
    5: "Solo",
    6: "Chamber",
    7: "Pop",
    8: "Dance_and_house",
    9: "Indie",
    10: "Soul_or_r_and_b",
    11: "Rock"
}

_NAMES_3 = {
    0: "None",
    12: "Pop_vocal_ballad",
    13: "Adult_contemporary",
    14: "Teen_pop",
    15: "Contemporary_dance_pop",
    16: "Dance_pop",
    17: "Classic_indie_pop",
    18: "Chamber_cabaret_and_art_pop",
    19: "Adult_alternative_rock",
    20: "Uplifting_anthemic_rock",
    21: "Soft_rock",
    22: "Acoustic_pop"
}

_HOMEPAGE = f"https://huggingface.co/datasets/ccmusic-database/{os.path.basename(__file__).split('.')[0]}"

_CITATION = """\
@dataset{zhaorui_liu_2021_5676893,
  author       = {Zhaorui Liu, Monan Zhou, Shenyang Xu, 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}
}
"""

_DESCRIPTION = """\
This database contains about 1700 musical pieces (.mp3 format) 
with lengths of 270-300s that are divided into 17 genres in total. 
"""

_URL = _HOMEPAGE + "/resolve/main/data/mel.zip"


class music_genre(datasets.GeneratorBasedBuilder):
    def _info(self):
        return datasets.DatasetInfo(
            features=datasets.Features(
                {
                    "mel": datasets.Image(),
                    "fst_level_label": datasets.features.ClassLabel(names=list(_NAMES_1.values())),
                    "sec_level_label": datasets.features.ClassLabel(names=list(_NAMES_2.values())),
                    "thr_level_label": datasets.features.ClassLabel(names=list(_NAMES_3.values()))
                }
            ),
            supervised_keys=("mel", "sec_level_label"),
            homepage=_HOMEPAGE,
            license="mit",
            citation=_CITATION,
            description=_DESCRIPTION,
            task_templates=[
                ImageClassification(
                    task="image-classification",
                    image_column="mel",
                    label_column="sec_level_label",
                )
            ]
        )

    def _split_generators(self, dl_manager):
        data_files = dl_manager.download_and_extract(_URL)
        files = dl_manager.iter_files([data_files])

        dataset = []
        for path in files:
            if os.path.basename(path).endswith(".jpg"):
                dataset.append(path)

        random.shuffle(dataset)
        data_count = len(dataset)
        p80 = int(data_count * 0.8)
        p90 = int(data_count * 0.9)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "files": dataset[:p80]
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                gen_kwargs={
                    "files": dataset[p80:p90]
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={
                    "files": dataset[p90:]
                },
            ),
        ]

    def _calc_label(self, path, depth, substr='/mel/'):
        mel = substr
        dirpath = os.path.dirname(path)
        substr_index = dirpath.find(mel)
        if substr_index < 0:
            mel = '\\mel\\'
            substr_index = dirpath.find(mel)

        labstr = dirpath[substr_index + len(mel):]
        labs = labstr.split('/')
        if len(labs) < 2:
            labs = labstr.split('\\')

        if depth <= len(labs):
            return 0
        else:
            return int(labs[-1].split('_')[0])

    def _generate_examples(self, files):
        for i, path in enumerate(files):
            yield i, {
                "mel": path,
                "fst_level_label": _NAMES_1[self._calc_label(path, 1)],
                "sec_level_label": _NAMES_2[self._calc_label(path, 2)],
                "thr_level_label": _NAMES_3[self._calc_label(path, 3)]
            }