数据堂—13.8小时中文女声情感合成库

13.8小时中文女声情感合成库，由中文母语发音人录制，六种情感文本，语料音素覆盖均衡，专业语音学家参与标注，精准匹配语音合成的研发需求

This 13.8-hour Mandarin female emotional speech synthesis corpus was recorded by native Mandarin-speaking female speakers. It encompasses textual resources across six distinct emotional categories, boasts balanced phoneme coverage across the entire dataset, and was annotated by professional speech phoneticians, making it perfectly aligned with the research and development requirements of speech synthesis systems.

搜集汇总

背景与挑战

背景概述

该数据集是一个13.8小时的中文女声情感语音合成库，专为语音合成测试任务设计。它由母语中文女性录制，涵盖六种情感类型，并经过专业标注，采用48kHz、16bit的wav格式，适用于商业用途。