Western Yugur Two-Speaker Parallel Lexical Speech Dataset

During the sorting process, consistent terms from the two language information templates were used as a unified index, and three vocabulary and annotation tables were cross checked. The final WAV directory was used to verify the numbering, unified term names, actual file names, relative paths, duration, and audio technology parameters one by one. The verification results indicate that the word item numbers of the two speakers continuously cover 1-3070, without missing numbers, duplicate numbers, or out of range numbers; 6140 WAV files can be read normally, and the file name, relative path, actual duration, and technical parameters are consistent with the final metadata one by one.

This dataset can provide basic data for the preservation of Western Yugur vocabulary phonetics, cross pronunciation comparison under unified word items, manual listening and subsequent transcription annotation, as well as digital organization of ethnic language resources.