当前位置: X-MOL 学术IEEE ACM Trans. Audio Speech Lang. Process. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
The Harmonic Shift Algorithm for Efficient Multi-Pitch Detection
IEEE/ACM Transactions on Audio, Speech, and Language Processing ( IF 4.1 ) Pub Date : 2021-11-19 , DOI: 10.1109/taslp.2021.3129344
Lukas Grinewitschus , Peter Jung

Existing approaches for multi-pitch detection have still been highly complex. In this paper, the harmonic shift algorithm (HSA) for multi-pitch detection is illustrated. The HSA is a noncoherent blind detection scheme, exploiting the constant-Q transform (CQT). The only requirement for the successful application of the HSA is that all the sources are harmonic sources. Due to this simplicity, the HSA is considered viable for both single-pitch and multi-pitch detection, making its attractive for the application in other areas of signal processing. When illustrating the performance capabilities of the HSA, we demonstrate that it outperforms the state of the art regarding the computational complexity providing detection precision, accuracy and a recall factor comparable with existing schemes.

中文翻译:


用于高效多音高检测的谐波移位算法



现有的多音高检测方法仍然非常复杂。本文阐述了用于多基音检测的谐波移位算法(HSA)。 HSA 是一种非相干盲检测方案,利用恒定 Q 变换 (CQT)。 HSA成功应用的唯一要求是所有源都是谐波源。由于这种简单性,HSA 被认为适用于单节距和多节距检测,这使其对于其他信号处理领域的应用具有吸引力。在说明 HSA 的性能时,我们证明它在计算复杂性方面优于现有技术,提供与现有方案相当的检测精度、准确度和召回因子。
更新日期:2021-11-19
down
wechat
bug