paddlespeech.t2s.exps.ernie_sat.align module

paddlespeech.t2s.exps.ernie_sat.align.alignment(wav_path: str, text: str, fs: int = 24000, lang='en', n_shift: int = 300)

paddlespeech.t2s.exps.ernie_sat.align.get_phns_spans(wav_path: str, old_str: str = '', new_str: str = '', source_lang: str = 'en', target_lang: str = 'en', fs: int = 24000, n_shift: int = 300)

paddlespeech.t2s.exps.ernie_sat.align.words2phns(text: str, lang='en')

Args:

text (str):

input text. eg: for that reason cover is impossible to be given.

lang (str):

'en' or 'zh'

Returns:

List[str]: phones of input text.

eg: ['F', 'AO1', 'R', 'DH', 'AE1', 'T', 'R', 'IY1', 'Z', 'AH0', 'N', 'K', 'AH1', 'V', 'ER0', 'IH1', 'Z', 'IH2', 'M', 'P', 'AA1', 'S', 'AH0', 'B', 'AH0', 'L', 'T', 'UW1', 'B', 'IY1', 'G', 'IH1', 'V', 'AH0', 'N']

Dict(str, str): key - idx_word

value - phones

eg: {'0_FOR': ['F', 'AO1', 'R'], '1_THAT': ['DH', 'AE1', 'T'], '2_REASON': ['R', 'IY1', 'Z', 'AH0', 'N'],'3_COVER': ['K', 'AH1', 'V', 'ER0'], '4_IS': ['IH1', 'Z'], '5_IMPOSSIBLE': ['IH2', 'M', 'P', 'AA1', 'S', 'AH0', 'B', 'AH0', 'L'], '6_TO': ['T', 'UW1'], '7_BE': ['B', 'IY1'], '8_GIVEN': ['G', 'IH1', 'V', 'AH0', 'N']}