A multi layered signal enhancement and machine learning model for discriminating between gravitational wave signals from core collapse supernovae.

Core collapse supernovae (CCSN) are highly anticipated sources of gravitational waves during the fourth observation run (O4). CCSN signals are weak and unmodeled and the rate of occurrence in our galaxy is very low. Because of this, they provide a greater challenge to detect than previously detected GW sources. CCSN simulations are used to test the detection pipeline in the event a CCSN is detected. CCSN GW signals are often indistinguishable from the noise sources present in GW data. We present a multi layered signal enhancement pipeline which we have applied Machine Learning (ML) techniques. We have used a Convolutional Neural Network (CNN) to train both the CCSN with simulated signals from Müller 2018, 3D simulations using a 39 solar mass progenitor, and the background noise. We demonstrate the results using O3 data in a two-detector network.