Research on automatic segmentation method of microseismic monitoring signal under low signal-to-noise ratio conditions
Weijian Liu, Jianbo Li, Zhizeng Zhang, Zhenxia Yuan, Shuai Teng, Danqing Song, Mingrui Zhu, Wenlong Lv · 2026
Read Research on automatic segmentation method of microseismic monitoring signal under low signal-to-noise ratio conditions by Weijian Liu, Jianbo Li, Zhizeng Zhang, Zhenxia Yuan, Shuai Teng, Danqing Song, Mingrui Zhu, Wenlong Lv (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0358562.txt.
Description
Weijian Liu leads a team of eight building software to pick tremor events out of very noisy seismic recordings, in PLOS ONE in 2026. Microseismic monitoring listens for tiny rock fractures in mines and reservoirs, but under heavy noise the start and end of each event blur. Their model reshapes a standard segmentation network with an attention module that reweights signal channels toward event boundaries, and stretches one dimensional waveforms into a form two dimensional networks can read.
Validated on acoustic emission data from rock compression tests, the system segmented events with accuracy near ninety-nine percent, about a point above its parent network. Under strong noise, four fifths of start picks and nine tenths of end picks landed within a tight tolerance. It also outscored a recognized rival method on both detection quality measures reported. Reliable boundaries are what turn raw rumble into usable warnings about rock under stress.
Catalog record
| CREATOR | Weijian Liu, Jianbo Li, Zhizeng Zhang, Zhenxia Yuan, Shuai Teng, Danqing Song, Mingrui Zhu, Wenlong Lv |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43986 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |