AAdonis/multilingual_audio_alignments
收藏资源简介:
这是一个大规模的多语言语音数据集,使用蒙特利尔强制对齐器(MFA)提供了精确的单词级别和音素级别的对齐信息。数据集整合了多种语言的语音语料库,每个样本包括原始音频、转录文本以及单词和音素的详细时间信息。数据集支持多种语言,包括英语、德语、法语、西班牙语、俄语、日语、韩语、葡萄牙语、土耳其语和泰语等。总时长估计超过20,000小时。数据集的主要特征包括音频波形、文本转录、音素序列、单词和音素的对齐信息以及数据来源。
A large-scale multilingual speech dataset with word-level and phoneme-level alignments produced using the Montreal Forced Aligner (MFA). This dataset consolidates multiple speech corpora across various languages, all processed through MFA to provide precise phoneme and word alignments. Each sample includes the original audio, transcript, and detailed timing information for both words and phonemes. The dataset supports multiple languages including English, German, French, Spanish, Russian, Japanese, Korean, Portuguese, Turkish, and Thai, with an estimated total duration of over 20,000 hours. Key features include audio waveforms, text transcripts, phoneme sequences, word and phoneme alignments, and data sources.




