FatimahEmadEldin/Arabic-Emotional-Audio-Dataset-Baved
收藏资源简介:
BAVED(基本阿拉伯语语音情感数据集)是一个重新打包的、转录对齐的阿拉伯语语音情感数据集,包含明确的阿拉伯语转录、英语注释、说话者元数据以及说话者不相交的训练/验证/测试分割。数据集包含1935个独特的录音,来自60个说话者(44名男性,16名女性),年龄在5至41岁之间。涵盖了7个阿拉伯语单词和3种情感强度级别,共21种独特的文本-情感对。录音以16 kHz单声道WAV格式提供。数据集适用于语音情感识别、情感/副语言TTS研究以及带情感调节的关键词识别。
BAVED — Basic Arabic Vocal Emotions Dataset (TTS-ready repackaging) is a re-packaged, transcript-aligned version of the Basic Arabic Vocal Emotions Dataset (BAVED) with explicit Arabic transcripts, English glosses, speaker metadata, and speaker-disjoint train/validation/test splits. It includes 1935 unique recordings from 60 speakers (44 male, 16 female), aged 5–41. The dataset covers 7 Arabic words and 3 emotion intensity levels, resulting in 21 unique text-emotion pairs. Recordings are provided in 16 kHz mono WAV format. The dataset is useful for speech-emotion recognition, emotional/paralinguistic TTS research, and keyword spotting with affective conditioning.




