12小时中文女声娱乐主播风格多情感合成库【数据堂】

12小时娱乐主播风格多情感合成库，由中文母语发音人录制。七种情感数据+语气词。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 12-hour multi-emotional speech synthesis dataset is styled in the manner of entertainment anchors, recorded by native Mandarin speakers. It includes data for seven distinct emotional states and interjections, with balanced phoneme coverage across the entire corpus. Annotated by professional phoneticians, this dataset accurately meets the research and development requirements of speech synthesis.

搜集汇总

背景与挑战

背景概述

该数据集是一个12小时的中文女声娱乐主播风格多情感合成音频库，由专业声优录制，涵盖七种情感和语气词，采用专业录音棚环境和精细标注，适用于语音合成研发。