eduhk-compling/11457167_JaanAlyana

该数据集包含51个乌尔都语句子级别的音频记录，配有乌尔都语文本和罗马化乌尔都语转录，总计超过3分钟的语音。句子内容涵盖日常话题，如日常生活、情感和爱好，句子长度、结构和语音模式多样，旨在反映自然的口语表达。所有录音由母语者在安静室内环境中使用手机麦克风录制，音频剪辑为16位PCM WAV格式，采样率为44.1kHz，并配有结构化的metadata.csv文件以便机器读取。

This dataset contains 51 sentence-level audio recordings in Urdu with both Urdu text and Roman Urdu transcriptions, totalling over 3 minutes of speech. The sentences focus on everyday topics such as daily routines, emotions, and hobbies, with variation in sentence length, structure, and phonetic patterns, aiming to reflect natural, everyday spoken language. All recordings were produced by a native speaker in a quiet indoor environment using a phone microphone. The audio clips are saved in 16-bit PCM WAV format at 44.1kHz and are paired with corresponding transcriptions in a structured metadata.csv for machine-readable access.