Most current searches for gravitational waves in data from the interferometers deploy matched filtering techniques. However, the time complexity of these searches directly scales with the number of templates and their duration. As these interferometers approach their design sensitivity and with the advent of new models to probe into a more expansive parameter space, the time complexity of these searches is bound to increase tremendously as the rate of occurrence of glitches will shoot up with the non-zero probability of discovering new types of glitches to hinder our searches.
Deep learning is a possible way to make these searches low latency by leveraging the fact that the model has to be trained only once, and can then be deployed online, orders of magnitudes faster than the current search algorithms.
We have developed a robust Deep Transfer Learning model to detect Intermediate Mass Black Hole (IMBH) binaries. This model is based on an Inception-v3 architecture. Detailed results from this project were presented at ASI 2022 Conference.