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 2104.03538cs.SDcs.AIeess.AS
收藏

MetricGAN +:一种改进的MetricGAN语音增强算法

MetricGAN+: An Improved Version of MetricGAN for Speech Enhancement

Research Center for Information Technology Innovation, Academia Sinica(中央研究院资讯科技创新研究中心);The Ohio State University(俄亥俄州立大学);Mila-Quebec AI Institute(米拉-魁北克人工智能研究所);National Institute of Information and Communications Technology(日本信息通信研究机构)

Szu-Wei Fu, Cheng Yu, Tsun-An Hsieh, Peter Plantinga, Mirco Ravanelli, Xugang Lu, Yu Tsao

摘要:

用于训练语音增强模型的代价函数与人类听觉感知之间的差异通常使得增强后的语音质量不令人满意。因此,考虑人类感知的客观评价指标可以作为缩小差距的桥梁。我们之前提出的MetricGAN旨在通过将度量与一个参数连接来优化客观度量。由于在训练过程中只需要目标评价函数的分数,因此度量甚至可以是不可微的。在这项研究中,我们提出了一个MetricGAN+,其中提出了三个训练技术,将语音处理的领域知识。通过这些技术,在VoiceBank-DEMAND数据集上的实验结果表明,与之前的MetricGAN相比,MetricGAN+可以将PESQ得分提高0.3,并实现最先进的结果(PESQ得分= 3.15)。

英文摘要:

The discrepancy between the cost function used for training a speech enhancement model and human auditory perception usually makes the quality of enhanced speech unsatisfactory. Objective evaluation metrics which consider human perception can hence serve as a bridge to reduce the gap. Our previously proposed MetricGAN was designed to optimize objective metrics by connecting the metric with a discriminator. Because only the scores of the target evaluation functions are needed during training, the metrics can even be non-differentiable. In this study, we propose a MetricGAN+ in which three training techniques incorporating domain-knowledge of speech processing are proposed. With these techniques, experimental results on the VoiceBank-DEMAND dataset show that MetricGAN+ can increase PESQ score by 0.3 compared to the previous MetricGAN and achieve state-of-the-art results (PESQ score = 3.15).

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。

↑