50人中英混读客服平均音色合成库【数据堂】

50人中英混读客服平均音色合成库，由中文母语发音人录制，客服场景的录音文本，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。共计28.7小时

This is a 50-speaker Mandarin-English code-switching customer service speech synthesis corpus with averaged timbre. Recorded by native Mandarin speakers, it uses scripted texts tailored for customer service scenarios, features balanced phoneme coverage across the entire corpus, and was professionally annotated by phoneticians to precisely meet the R&D requirements of speech synthesis systems, with a total duration of 28.7 hours.

搜集汇总

背景与挑战

背景概述

该数据集为50名20-30岁中文母语发音人录制的客服场景中英混读语音库，总时长28.7小时，音频格式为48kHz、16bit、wav单声道，覆盖均衡的音素和韵律标注，专为语音合成研发设计，但属于需企业合作购买的商业数据。