CNN-Based Classification of Gravitational-Wave Bursts and Glitches

Distinguishing astrophysical GWBs from instrumental glitches remains a critical challenge for ground-based detectorslike LIGO, Virgo, and KAGRA. Among the most prolific offenders are blip glitches, whose short-duration, broadbandmorphology closely mimics unmodeled burst signals such as CSG waveforms. This morphological similarity complicatesreal-time detection pipelines and increases false alarm rates near the detection threshold. We present a 1D CNN designedfor binary classification between simulated CSG bursts and blips. At the LIGO detection threshold of SNR = 8, whereclassification is most challenging, the model achieves 96.9% accuracy and an F1-score of 0.968. Performance improvesmonotonically with signal strength, reaching >99% accuracy at SNR = 24 and SNR = 32.4. These results demonstrate thatcomputationally efficient 1D architectures can provide robust glitch classification and rejection for low-latency pipelines,with performance suitable for integration into future observing runs.