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 2410.17834eess.AScs.LGcs.SD
收藏

基于仅用干净语音训练的扩散模型的非侵入式语音质量评估

Non-intrusive Speech Quality Assessment with Diffusion Models Trained on Clean Speech

University of Hamburg(汉堡大学)

Danilo de Oliveira, Julius Richter, Jean-Marie Lemercier, Simon Welker, Timo Gerkmann

AI总结:

本研究探索扩散模型在语音密度估计的潜力,提出仅用干净语音训练的无监督非侵入式语音质量评估方法,通过确定性加噪得到的终端高斯分布似然评估质量,与侵入式指标及人类评分相关性良好。

摘要:

扩散模型在生成高质量、自然的语音样本方面已取得巨大成功,但迄今为止,其在语音密度估计方面的潜力在很大程度上仍未被探索。在本研究中,我们利用一个仅在干净语音上训练的无条件扩散模型来进行语音质量评估。我们表明,一段语音的质量可以通过估计其在终端高斯分布中对应样本的似然来评估,该样本通过确定性加噪过程获得。所提出的方法完全是无监督的,仅在干净语音上训练,因此不依赖标注。我们基于扩散的方法利用干净语音先验,根据输入与学习到的干净数据分布的关联程度来评估质量。我们提出的对数似然值展现出了可观的结果,与侵入式语音质量指标具有良好的相关性,并且在一项听力实验中与人类评分的相关性达到最佳。

英文摘要:

Diffusion models have found great success in generating high quality, natural samples of speech, but their potential for density estimation for speech has so far remained largely unexplored. In this work, we leverage an unconditional diffusion model trained only on clean speech for the assessment of speech quality. We show that the quality of a speech utterance can be assessed by estimating the likelihood of a corresponding sample in the terminating Gaussian distribution, obtained via a deterministic noising process. The resulting method is purely unsupervised, trained only on clean speech, and therefore does not rely on annotations. Our diffusion-based approach leverages clean speech priors to assess quality based on how the input relates to the learned distribution of clean data. Our proposed log-likelihoods show promising results, correlating well with intrusive speech quality metrics and showing the best correlation with human scores in a listening experiment.

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。

↑