arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.05575cs.SDeess.AS
收藏

SB-RF: 用于一步鲁棒语音增强的薛定谔桥整流流

SB-RF: Schrödinger Bridge Rectified Flow for One-Step Robust Speech Enhancement

Xiaomi Corporation, Beijing, China(小米公司,北京,中国)

Caixia Lu, Xueyang Lv, Penglong Hu, Jiaming Xu

AI总结:

提出SB-RF,一种结合整流流与薛定谔桥理论的一步生成式语音增强框架,通过熵正则化最优传输构建条件桥,实现高质量一步生成,在VoiceBank-DEMAND基准上达到领先性能,并在低信噪比场景下展现出强鲁棒性和高效率。

摘要:

生成模型在语音增强中表现出令人印象深刻的结果,但通常受限于多步推理。我们提出SB-RF,一种将整流流(RF)与薛定谔桥(SB)理论相结合的一步生成框架。SB-RF通过熵正则化最优传输在干净和带噪语音分布之间构建条件桥。通过RF的速度匹配目标将SB轨迹与最优传输测地线对齐,SB-RF能够通过一步生成实现高质量增强。实验表明,SB-RF在VoiceBank-DEMAND基准上达到了生成方法中的领先性能。此外,为了全面评估在具有挑战性的真实场景中的性能,我们在一个模拟的低信噪比测试集上使用扩大的训练数据集评估SB-RF。在这些条件下,SB-RF展现出强大且具有竞争力的鲁棒性和高效率,验证了其在现实应用中的潜力。

英文摘要:

Generative models have shown promising results for speech enhancement (SE), but they often rely on multi-step inference, limiting low-latency deployment. We propose SB-RF, a one-step generative framework that integrates Rectified Flow (RF) with Schrödinger Bridge (SB) theory. During training, SB-RF samples intermediate states from an SB time marginal and trains a conditional velocity field with the RF velocity-matching objective. At inference, SB-RF starts from the noisy observation and applies a single Euler update. Experiments show that SB-RF achieves competitive performance among generative methods on the VoiceBank-DEMAND benchmark. To further assess performance beyond this standard setting, we evaluate SB-RF on a simulated low signal-to-noise ratio test set using an expanded training dataset. Under these conditions, SB-RF achieves superior performance over the compared baselines, supporting its potential for real-world applications.

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。

↑