Development of a new pipeline leveraging convolutional neural networks for the detection of gravitational waves in signals from the VIRGO project's interferometer. The internship aimed to improve upon the existing statistical method, which involved an intensive process of correlating the entire interferometer signal with a broad array of signatures from various types of binary astrophysical system collapses. This traditional method is resource intensive. My development focused on a MVP that demonstrated the feasibility of using AI to preprocess the interferometer signal. The objective was to substantially narrow down the range of the signal processed by the statistical method, while still maintaining recall close to 100%.

Développement de le première version du système de recommandation de la marketplace de petites annonces