arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.08745eess.AScs.SD
收藏

More for Less:有限注释的非侵入式语音质量评估

More for Less: Non-Intrusive Speech Quality Assessment with Limited Annotations

University College Dublin(都柏林大学学院);Insight Centre for Data Analytics(洞察数据分析中心);Queen Mary University of London(伦敦玛丽女王大学);The Alan Turing Institute(艾伦·图灵研究所)

Alessandro Ragano, Emmanouil Benetos, Andrew Hines

摘要:

非侵入式语音质量评估是多媒体应用中的一项关键操作。注释数据的稀缺和缺乏参考信号是设计有效的质量评估指标的一些主要挑战。在本文中,我们提出了两个多任务模型来解决上述问题。在第一个模型中,我们首先在一个大型数据集上学习一个退化分类器的特征表示。然后,我们进行MOS预测和退化分类同时在一个小的数据集上标注MOS。在第二种方法中,初始阶段包括在大型数据集上使用基于深度聚类的无监督特征表示来学习特征。接下来,我们在一个小数据集上同时执行MOS预测和聚类标签分类。结果表明,基于深度聚类的模型在TCD-VoIP上的性能优于基于降级分类器的模型和3个基线(自动编码器特征、P.563和SRMRnorm)。本文指出,多任务学习与未标记数据的特征表示相结合是一种很有前途的方法,可以处理缺乏大型MOS注释数据集的问题。

英文摘要:

Non-intrusive speech quality assessment is a crucial operation in multimedia applications. The scarcity of annotated data and the lack of a reference signal represent some of the main challenges for designing efficient quality assessment metrics. In this paper, we propose two multi-task models to tackle the problems above. In the first model, we first learn a feature representation with a degradation classifier on a large dataset. Then we perform MOS prediction and degradation classification simultaneously on a small dataset annotated with MOS. In the second approach, the initial stage consists of learning features with a deep clustering-based unsupervised feature representation on the large dataset. Next, we perform MOS prediction and cluster label classification simultaneously on a small dataset. The results show that the deep clustering-based model outperforms the degradation classifier-based model and the 3 baselines (autoencoder features, P.563, and SRMRnorm) on TCD-VoIP. This paper indicates that multi-task learning combined with feature representations from unlabelled data is a promising approach to deal with the lack of large MOS annotated datasets.

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。

↑