Singing voice conversion method based on many-to-many eigenvoice conversion and training data generation using a singing-to-singing synthesis system
Singing voice conversion method based on many-to-many eigenvoice conversion and training data generation using a singing-to-singing synthesis system
复制标题
DOI：
--
复制
发表时间：
2012-12
期刊：
影响因子：
--
通讯作者：
Hironori Doi;T. Toda;Tomoyasu Nakano;Masataka Goto;Satoshi Nakamura
中科院分区：
文献类型：
--
作者：
Hironori Doi;T. Toda;Tomoyasu Nakano;Masataka Goto;Satoshi Nakamura
歌唱声音的声音质量（身份）通常在每个歌手中是固定的。为了克服这一限制，使歌手可以自由地改变自己的声音质量使用信号处理技术，我们提出了一种基于多对多特征声转换（EVC）的歌唱声音转换方法，可以转换成另一个任意的目标歌手的语音质量的任意源歌手。以前的基于EVC的方法需要由单个参考歌手和许多预存储的目标歌手的歌曲对组成的并行数据来训练语音转换模型，但是很难记录这样的数据。因此，我们所提出的方法使用一个唱歌唱歌的合成系统，称为Vocaetry，以产生并行数据，通过模仿唱歌的声音，许多预先存储的目标歌手与系统的歌声。实验结果表明，我们的方法成功地使人们能够唱一首歌曲的声音质量不同的目标歌手，即使只有一个非常少量的目标歌唱的声音是可用的。
The voice quality (identity) of singing voices is usually fixed in each singer. To overcome this limitation and enable singers to freely change their voice quality using signal-processing technologies, we propose a singing voice conversion method based on many-to-many eigenvoice conversion (EVC) that can convert the voice quality of an arbitrary source singer into that of another arbitrary target singer. Previous EVC-based methods required parallel data consisting of song pairs of a single reference singer and many prestored target singers for training a voice conversion model, but it was difficult to record such data. Our proposed method therefore uses a singing-to-singing synthesis system called VocaListener to generate parallel data by imitating singing voices of many prestored target singers with the system's singing voices. Experimental results show that our method succeeded in enabling people to sing a song with the voice quality of a different target singer even if only an extremely small amount of the target singing voice is available.