AI and the World's Hard Problems · Physics · arXiv · 2026-10-22
Deep Learning Search for Gravitational Waves from Compact Binary Coalescence
Matched filtering remains central to gravitational-wave searches for compact binary coalescences, but its computational cost will rise substantially with the data volumes and parameter-space coverage needed by next-generation interferometers such as the Einstein Telescope. The authors investigate a hybrid approach combining matched-filtering concepts with convolutional neural networks, enabling efficient signal searches without the usual χ² rejection test. On simulated data where glitches affect the data or the signals contain physics absent from the template bank—eccentricity, precession, and higher-order modes—the method achieves detection efficiency comparable to or better than a standard matched-filtering search with a more resource-efficient pipeline, while showing some loss of detection capability for standard BNS/BBH searches in pure Gaussian noise.