selfclean_audio.datasets.gtzan#

Members

GTZANKnownIssuesDataset

GTZAN dataset wrapper with built-in access to known data quality issues.

class selfclean_audio.datasets.gtzan.GTZANKnownIssuesDataset(root: str | Path, issue_type: str = 'duplicates', gt_duplicates_file: str | Path | None = None, gt_prep_file: str | Path | None = None, convert_mono: bool = True, sample_rate: int = 16000, target_duration_sec: float | None = 30.0, extensions: tuple[str, ...] = ('.wav', '.mp3', '.flac'))[source]#

GTZAN dataset wrapper with built-in access to known data quality issues.

  • Exposes audio samples from a local GTZAN folder (genres/<class>/*.wav)

  • Parses ground truth CSVs with known issues from external_code

  • Provides get_errors() compatible with SelfClean evaluation:
    • For ISSUE_TYPE “duplicates”: returns (set of (idx_i, idx_j) pairs, [..labels..])

    • For ISSUE_TYPE “label_errors”: returns a list[int] of 0/1 per sample

get_errors()[source]#

Return evaluation ground truth in the format expected by SelfClean.

  • If evaluating duplicates: returns (set[(i, j)], [..labels..])

  • If evaluating label errors: returns list[int] of length len(self)