paddleaudio.datasets.rirs_noises 模块

class paddleaudio.datasets.rirs_noises.OpenRIRNoise(subset: str = 'rir', feat_type: str = 'raw', target_dir=None, random_chunk: bool = True, chunk_duration: float = 3.0, seed: int = 0, **kwargs)[来源]

基础： Dataset

方法

META_INFO 的别名

生成_csv
准备数据

archieves = [{'url': 'http://www.openslr.org/resources/28/rirs_noises.zip', 'md5': 'e6f48e257286e05de56413b4779d8ffb'}]

base_path = '/home/docs/.paddleaudio/datasets/open_rir_noise'

csv_path = '/home/docs/.paddleaudio/datasets/open_rir_noise/csv'

generate_csv(wav_files: List[str], output_file: str, split_chunks: bool = True)[来源]

meta_info: 的别名 META_INFO

prepare_data()[来源]

sample_rate = 16000

subsets = ['rir', 'noise']

wav_path = '/home/docs/.paddleaudio/datasets/open_rir_noise/RIRS_NOISES'