arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.11981cs.SDcs.AIeess.AS
收藏

使用数据仿真的无监督噪声自适应

Unsupervised Noise adaptation using Data Simulation

Nanyang Technological University(南洋理工大学);Nanjing University of Posts and Telecommunications(南京邮电大学)

Chen Chen, Yuchen Hu, Heqing Zou, Linhui Sun, Eng Siong Chng

摘要:

基于深度神经网络的语音增强方法旨在使用监督学习范例来学习噪声到干净的变换。然而,这种训练良好的变换容易受到未包括在训练集中的不可见噪声的影响。在这项工作中,我们专注于语音增强中的无监督噪声自适应问题,其中目标域数据的地面真值完全不可用。具体来说,我们提出了一种基于生成对抗网络的方法,使用几分钟的未配对目标域数据来有效地学习逆清洁到噪声变换。然后,利用该变换来生成用于增强模型的领域自适应的足够的模拟数据。实验结果表明,该方法有效地缓解了训练集和测试集之间的领域不匹配,并大大超过了最佳基线。

英文摘要:

Deep neural network based speech enhancement approaches aim to learn a noisy-to-clean transformation using a supervised learning paradigm. However, such a trained-well transformation is vulnerable to unseen noises that are not included in training set. In this work, we focus on the unsupervised noise adaptation problem in speech enhancement, where the ground truth of target domain data is completely unavailable. Specifically, we propose a generative adversarial network based method to efficiently learn a converse clean-to-noisy transformation using a few minutes of unpaired target domain data. Then this transformation is utilized to generate sufficient simulated data for domain adaptation of the enhancement model. Experimental results show that our method effectively mitigates the domain mismatch between training and test sets, and surpasses the best baseline by a large margin.

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。

↑