KALLAAMA
收藏资源简介:
KALLAAMA项目旨在为塞内加尔最广泛使用的三种语言(Wolof, Pulaar, Sereer)创建语音技术所需的资源。该项目专注于开发能够通过语音查询设备访问数字信息的语音转文本(ASR)和文本转语音(TTS)模型。数据集主要关注农业主题,包含总计125小时的转录语音,其中包括35小时的专家验证内容。
The KALLAAMA project aims to create the necessary resources for speech technology in the three most widely used languages in Senegal (Wolof, Pulaar, Sereer). The project focuses on developing speech-to-text (ASR) and text-to-speech (TTS) models that can access digital information through voice queries. The dataset primarily focuses on agricultural topics, containing a total of 125 hours of transcribed speech, including 35 hours of expert-verified content.
项目概述
KALLAAMA项目旨在为塞内加尔最广泛使用的三种语言——Wolof、Pulaar和Sereer,开发语音技术所需的资源。该项目由Lacuna Fund资助,Jokalante领导,并得到Orange Innovation和Ecole Polytechnique de Thiès的支持。
数据集内容
- 语言与时长:
- Wolof (ISO Code 639-2: wol): 55小时转录语音,其中13小时经过专家验证。
- Pulaar (ISO Code 639-2: fuc): 32小时转录语音,其中11小时经过专家验证。
- Sereer (ISO Code 639-2: srr): 38小时转录语音,其中11小时经过专家验证。
- 总时长: 125小时转录语音,包括35小时验证过的转录。
- 内容主题: 农业相关。
- 数据类型: 包括口语录音、正字法转录、开源文本收集、词汇表及音标转录,以及为Wolof语言训练的音素到音素的模型。
数据集结构
-
目录结构:
. ├── LICENSE ├── README.md └── data/ ├── README.md ├── lexicons/ ├── text_corpora/ └── transcriptions/ ├── checked/ └── raw/
引用信息
-
出版物: Kallaama: A Transcribed Speech Dataset about Agriculture in the Three Most Widely Spoken Languages in Senegal
-
作者: Gauthier, Elodie; Ndiaye, Aminata; Guissé, Abdoulaye
-
会议: Proceedings of the Fifth workshop on Resources for African Indigenous Languages (RAIL 2024)
-
年份: 2024
-
引用格式:
@inproceedings{kallaama2024dataset, title={Kallaama: A Transcribed Speech Dataset about Agriculture in the Three Most Widely Spoken Languages in Senegal}, author={Gauthier, Elodie and Ndiaye, Aminata and Guissé, Abdoulaye}, booktitle={Proceedings of the Fifth workshop on Resources for African Indigenous Languages (RAIL 2024)}, year={2024} }




