scb10x/thaimos-tts-annotation

泰语TTS合成语音人类评估数据集，包含声音质量、静音部分和发音的人类评分，基于较旧的TTS模型开发于2024年，可能不适用于现代TTS系统。数据集由Datawow进行标注，每个样本由5到8个在泰国曼谷出生和长大并在安静环境中评估语音的听众进行评分。

Thai TTS Synthesized Speech Human Evaluation Dataset, containing human ratings on sound quality, silence, and pronunciation, based on older TTS models developed in 2024, possibly not applicable to modern TTS systems. The dataset was annotated by Datawow, with each sample rated by 5-8 listeners born and raised in Bangkok, Thailand, in a quiet environment.