arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.12937cs.SDeess.AS
收藏

SelfRemaster:基于信道建模的综合分析自监督语音恢复

SelfRemaster: Self-Supervised Speech Restoration with Analysis-by-Synthesis Approach Using Channel Modeling

The University of Tokyo(东京大学)

Takaaki Saeki, Shinnosuke Takamichi, Tomohiko Nakamura, Naoko Tanji, Hiroshi Saruwatari

摘要:

提出了一种无需配对语音库的自监督语音恢复方法。由于先前的一般语音恢复方法使用通过将各种失真应用于高质量语音语料库而创建的人工配对数据,因此其不能充分地表示真实数据的声学失真,从而限制了适用性。我们的模型包括分析,合成和通道模块,模拟退化语音的记录过程,并以自我监督的方式用真实的退化语音数据进行训练。分析模块从退化语音中提取无失真语音特征和失真特征,而合成模块合成恢复的语音波形,并且信道模块将失真添加到语音波形。我们的模型还支持音频效果传输,其中仅从降级的语音中提取声学失真并添加到任意高质量音频中。模拟和真实数据的实验评估表明,我们的方法实现了显着更高质量的语音恢复比以前的监督方法,表明其适用于真正的退化语音材料。

英文摘要:

We present a self-supervised speech restoration method without paired speech corpora. Because the previous general speech restoration method uses artificial paired data created by applying various distortions to high-quality speech corpora, it cannot sufficiently represent acoustic distortions of real data, limiting the applicability. Our model consists of analysis, synthesis, and channel modules that simulate the recording process of degraded speech and is trained with real degraded speech data in a self-supervised manner. The analysis module extracts distortionless speech features and distortion features from degraded speech, while the synthesis module synthesizes the restored speech waveform, and the channel module adds distortions to the speech waveform. Our model also enables audio effect transfer, in which only acoustic distortions are extracted from degraded speech and added to arbitrary high-quality audio. Experimental evaluations with both simulated and real data show that our method achieves significantly higher-quality speech restoration than the previous supervised method, suggesting its applicability to real degraded speech materials.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑