Nexdata/Saudi_Arabic_Spontaneous_Speech_Data
收藏Hugging Face2024-03-26 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/Nexdata/Saudi_Arabic_Spontaneous_Speech_Data
下载链接
链接失效反馈官方服务:
资源简介:
849小时的沙特阿拉伯自发语音数据,内容涵盖多个主题。所有语音音频都经过手动转录,并标注了说话者身份、性别等信息。该数据集可用于声纹识别模型训练、机器翻译语料库构建和算法研究。数据格式为16kHz、16bit、wav、单声道,内容包括访谈、综艺节目、直播等,语言为阿拉伯语。应用场景包括语音识别、视频字幕生成和视频内容审查,句子准确率不低于95%。
This is an 849-hour spontaneous speech dataset sourced from Saudi Arabia, covering a wide range of topics. All speech audio files have been manually transcribed, with annotations including speaker identity, gender, and other relevant metadata. This dataset can be used for voiceprint recognition model training, machine translation corpus construction, and algorithm research. The data is in WAV format with specifications of 16kHz sampling rate, 16-bit bit depth, and single-channel. The content includes interviews, variety shows, live broadcasts, and other types of materials, in Arabic. Its application scenarios include speech recognition, video subtitle generation, and video content moderation, with a sentence-level accuracy rate of no less than 95%.
提供机构:
Nexdata原始信息汇总
数据集概述
基本信息
- 名称: Nexdata/Saudi_Arabic_Spontaneous_Speech_Data
- 语言: 阿拉伯语
- 任务类别: 自动语音识别
数据集描述
- 时长: 849小时
- 内容: 涵盖多个话题的沙特阿拉伯自发语音数据
- 特征: 所有语音音频均手动转录为文本内容,包含说话人身份、性别等标注信息
- 应用: 用于声纹识别模型训练、机器翻译语料库构建及算法研究
规格
- 格式: 16kHz, 16bit, wav, 单声道
- 内容类别: 包括访谈、综艺节目、直播等
- 标注: 转录文本、说话人识别、性别标注
- 应用场景: 语音识别、视频字幕生成、视频内容审核
- 准确率: 句子准确率不低于95%
许可信息
- 许可类型: 商业许可



