KazEmoTTS

KazEmoTTS数据集由纳扎尔巴耶夫大学智能系统与人工智能研究所创建，专注于哈萨克语情感文本到语音合成。该数据集包含54,760个音频-文本对，总时长74.85小时，涵盖中性、愤怒、快乐、悲伤、害怕和惊讶六种情感。数据集由一名女性和两名男性专业叙述者录制，确保了情感表达的多样性。创建过程包括文本收集、录音和音频-文本对齐验证，旨在提高情感语音合成的质量和自然度，适用于语音情感识别和情感语音转换等领域。

The KazEmoTTS dataset was created by the Institute of Intelligent Systems and Artificial Intelligence at Nazarbayev University, focusing on Kazakh emotional text-to-speech synthesis. This dataset contains 54,760 audio-text pairs with a total duration of 74.85 hours, covering six emotions: neutral, angry, happy, sad, fearful and surprised. It was recorded by one female and two male professional narrators to ensure the diversity of emotional expressions. The creation process includes text collection, audio recording and audio-text alignment verification, aiming to improve the quality and naturalness of emotional speech synthesis, and is applicable to fields such as speech emotion recognition and emotional voice conversion.