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 2104.01227eess.AScs.SD
收藏

MetricNet:改进的非侵入式语音质量评估模型

MetricNet: Towards Improved Modeling For Non-Intrusive Speech Quality Assessment

Tencent AI Lab(腾讯人工智能实验室)

Meng Yu, Chunlei Zhang, Yong Xu, Shixiong Zhang, Dong Yu

摘要:

语音质量的客观评价通常是通过将接收到的语音信号与其干净的参考进行比较来进行的,而人类能够在没有任何参考的情况下评价语音质量,例如在平均意见得分(MOS)测试中。非侵入式语音质量评估最近引起了人们的广泛关注,由于缺乏获得干净的参考信号,在现实情况下的客观评价。在本文中,我们提出了一种新的非侵入式语音质量测量模型,MetricNet,它利用标签分布学习和联合语音重建学习,以实现显着提高性能相比,现有的非侵入式语音质量测量模型。我们证明,该方法产生了有希望的高相关性的侵入性客观评价的语音质量的干净,嘈杂和处理的语音数据。

英文摘要:

The objective speech quality assessment is usually conducted by comparing received speech signal with its clean reference, while human beings are capable of evaluating the speech quality without any reference, such as in the mean opinion score (MOS) tests. Non-intrusive speech quality assessment has attracted much attention recently due to the lack of access to clean reference signals for objective evaluations in real scenarios. In this paper, we propose a novel non-intrusive speech quality measurement model, MetricNet, which leverages label distribution learning and joint speech reconstruction learning to achieve significantly improved performance compared to the existing non-intrusive speech quality measurement models. We demonstrate that the proposed approach yields promisingly high correlation to the intrusive objective evaluation of speech quality on clean, noisy and processed speech data.

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。

↑