VoxClamantis V1.0

VoxClamantis V1.0是由约翰斯·霍普金斯大学等机构创建的第一个大规模语音类型学语料库，包含635种语言的690个读数，提供音段对齐和估计的音位级标签，以及元音和咝音的声学语音测量。该数据集旨在促进大规模跨语言语音类型学研究，解决跨语言数据不足的问题。数据集创建过程中采用了多策略强制对齐方法，确保广泛的语言覆盖和高质量资源利用。VoxClamantis V1.0适用于探索语音系统的普遍趋势和语音结构，为语音类型学研究提供了重要的数据支持。

VoxClamantis V1.0 is the first large-scale speech typology corpus compiled by institutions including Johns Hopkins University. It encompasses 690 recordings spanning 635 languages, providing segment-level alignment and estimated phonemic-level labels, as well as acoustic-phonetic measurements for vowels and sibilants. This dataset aims to facilitate large-scale cross-linguistic speech typology research and address the scarcity of cross-linguistic speech data. A multi-strategy forced alignment approach was adopted during the corpus construction process to ensure extensive language coverage and efficient utilization of high-quality resources. VoxClamantis V1.0 is suitable for exploring universal trends and phonetic structures of speech systems, offering critical data support for speech typology studies.