Mikimi/phoible-voxangeles-formants-audio-2
收藏资源简介:
该数据集是一个多语言音素声学特征数据库,包含来自多个语言的音素实现数据,用于语音学和音系学研究。数据集提供了详细的音素标注,包括国际音标(IPA)表示、音素基础形式、以及PHOIBLE数据库中的音素IPA对应。每个音素条目关联语言信息(如语言代码、语言名称、语言家族、ISO 639-3代码、地理位置和宏观区域)和声学特征,如基频(F0)及其分位数、共振峰(F1-F4)及其分位数、共振峰距离(赫兹和巴克尺度)。此外,数据集还包括音系特征标注,基于特征几何理论(如辅音性、鼻音性、圆唇性等),用于描述音素的语音属性。数据来源于多个语料库和录音,涵盖音素在单词中的上下文信息(如前/后音素、单词边界、时间戳)。数据集可用于音变分析、语音合成、语言比较研究等任务。
This dataset is a multilingual phoneme acoustic feature database containing phoneme realization data from multiple languages, intended for phonetics and phonology research. The dataset provides detailed phoneme annotations, including International Phonetic Alphabet (IPA) representations, basic forms of phonemes, and IPA correspondences of phonemes from the PHOIBLE database. Each phoneme entry is associated with language information such as language code, language name, language family, ISO 639-3 code, geographic location, and macro-area, as well as acoustic features including fundamental frequency (F0) and its quantiles, formants (F1-F4) and their quantiles, and formant distances measured in Hertz and Bark scale. In addition, the dataset includes phonological feature annotations based on feature geometry theory (e.g., consonantality, nasality, roundness, etc.) to describe the phonetic attributes of phonemes. The data is sourced from multiple corpora and audio recordings, covering contextual information of phonemes within words such as preceding and following phonemes, word boundaries, and timestamps. This dataset can be applied to tasks including sound change analysis, speech synthesis, and cross-linguistic comparative research.




