paddleaudio.datasets.gtzan module

class paddleaudio.datasets.gtzan.GTZAN(mode='train', seed=0, n_folds=5, split=1, feat_type='raw', **kwargs)[source]

Bases: AudioClassificationDataset

The GTZAN dataset consists of 1000 audio tracks each 30 seconds long. It contains 10 genres, each represented by 100 tracks. The dataset is the most-used public dataset for evaluation in machine listening research for music genre recognition (MGR).

Reference:: Musical genre classification of audio signals https://ieeexplore.ieee.org/document/1021072/

Methods

meta_info

alias of META_INFO

archieves = [{'url': 'http://opihi.cs.uvic.ca/sound/genres.tar.gz', 'md5': '5b3d6dddb579ab49814ab86dba69e7c7'}]

audio_path = 'genres'

label_list = ['blues', 'classical', 'country', 'disco', 'hiphop', 'jazz', 'metal', 'pop', 'reggae', 'rock']

meta = 'genres/input.mf'

meta_info: alias of META_INFO