Work

ayzek: a self-contained real-time earthquake detection and early-warning pipeline

Year: 2026

Type: software

Abstract: A C++23 earthquake detection and early-warning pipeline with hand-written inference and no machine-learning runtime: miniSEED ingestion, a learned 6 s window detector (with a STA/LTA reference trigger and a streaming transformer), P/S picking, network association, location and magnitude, replayed against AFAD and KOERI data. Includes the model weights, the tools that reproduce every number in the accompanying paper, and the benchmarks.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.23672495

Authors: Bolat, H. Oğuz, Çimen, Sibel, Bolat, Bülent, Yıldırım, Tülay

Institution: Yıldız Technical University

Language: English

Topic: Seismology and Earthquake Studies

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

OpenAlex ID: W7223657800

Cites:

Cited by:

Related to: 10