Umbaji/umini_speech

Umini_Speech是一个由Umbaji团队开发的迷你语音数据集，包含超过800个增强的Ewè语言单词录音，这些录音由不同用户在不同环境下录制。数据集设计用于语音识别任务，特别是Yodi模型。数据集包括WAV格式的音频，总时长为3.91小时，涵盖了多个单词类别及其对应的英文标签。数据集主要用于训练Yodi语音识别模型，也可以用于Ewè语言的语音识别任务。

Umini_Speech is a mini_speech dataset developed by the Umbaji team. It consists of over 800 augmented recordings of Ewè words, spoken by various users in different settings. The dataset is designed for speech recognition tasks, particularly for the Yodi model. It includes WAV format audio with a total duration of 3.91 hours, covering multiple word classes and their corresponding English labels. The dataset is primarily used to train the Yodi speech recognition model and can also be used for Ewè speech recognition tasks.

原始信息汇总

数据集许可证信息

• 许可证类型: CC0-1.0