AI Article Synopsis
- Conventional reverberation reduction methods often struggle in environments with low signal-to-reverberation ratios (SRR) because they only use single-ping data.
- This paper introduces an improved method that utilizes multi-ping data to effectively reduce both steady reverberation components and their fluctuations.
- The new approach showed better results when tested with real field data, achieving a 1.23 times improvement in sparse coefficients and an average 2.0 dB increase in signal quality, with performance linked more to the signal-to-reverberation-fluctuation ratio than just the SRR.
Article Abstract
Conventional reverberation reduction methods are conducted with single-ping data and may fail in a low signal-to-reverberation ratio (SRR) environment. To improve the performance of reverberation reduction, multi-ping data are fully considered in this paper. The reverberation can be treated as a combination of the steady component of reverberation and reverberation fluctuations, and then an alternating direction multiplier method is proposed to reduce the steady component of the reverberation. By exploiting the evolution of the target location along multiple pings, the reverberation fluctuation is reduced by the probabilistic data association method. The proposed method was verified by the field data, and the results show that compared with the accelerated proximal gradient method, the sparse coefficient is improved by a factor of 1.23, and the signal excess is improved by an average value of 2.0 dB. In addition, the performance of the proposed method is found to be closely related to the signal-to-reverberation-fluctuation ratio rather than only the SRR.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1121/10.0002175 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!