SPECTRA

The SPECTRA project aims to innovate astronomical observation by applying Artificial Intelligence (AI) to identify and analyze transient astronomical sources from high-energy gamma-ray data. Transient sources, such as gamma-ray bursts, supernovae and solar flares, are crucial phenomena to understand the high-energy Universe. However, the detection of such phenomena is challenging due to their ephemeral nature and the large amount of data produced by gamma-ray experiments. The approach proposed in this project aims to use Deep Learning algorithms to process and analyze spatial and temporal data from telescopes and satellites. In particular, complex Deep Learning architectures, such as Convolutional Neural Networks (CNNs) for image analysis and Recurrent Neural Networks (RNNs), including LSTMs, and Attention-based models such as Transformers, will be employed to handle photon emission time series and GNNs to analyze and model spatial relationships between various astronomical objects/events or recognize and classify complex patterns in the data that are not immediately evident through more traditional approaches.