42人中文多情感合成库【数据堂】

42人中文多情感合成库，由中文母语发音人录制，覆盖不同年龄段音色和性别。七种情感数据。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。共计108小时

42-speaker Chinese multi-emotional speech synthesis corpus. Recorded by native Chinese speakers, it covers timbres and genders across different age groups, and includes data for seven distinct emotions. The corpus has balanced phoneme coverage in its speech materials, with annotations conducted by professional phoneticians, which precisely meets the R&D requirements for speech synthesis. The total duration of the corpus is 108 hours.

搜集汇总

背景与挑战

背景概述

该数据集是一个由42名中文母语发音人录制的多情感语音合成库，覆盖不同年龄段和性别，包含高兴、愤怒、悲哀、惊讶、恐惧、厌恶和中性七种情感句子，总时长108小时，采用专业录音环境和标注，专为语音合成研发设计。作为商业数据，它仅面向企业合作提供。