A study of Neural networks point source extraction on simulated Fermi/LAT Telescope images
This work addresses a domain-specific problem for astrophysicists analyzing Fermi/LAT data, offering incremental improvements in accuracy and speed for point source extraction.
The paper tackled the problem of detecting point sources in challenging GeV-band astrophysical images from the Fermi/LAT telescope, where background noise and instrument limitations make extraction difficult. They developed a CNN-based method trained on simulated data, achieving a ~15% accuracy increase and at least 4x faster inference time compared to similar state-of-the-art models.
Astrophysical images in the GeV band are challenging to analyze due to the strong contribution of the background and foreground astrophysical diffuse emission and relatively broad point spread function of modern space-based instruments. In certain cases, even finding of point sources on the image becomes a non-trivial task. We present a method for point sources extraction using a convolution neural network (CNN) trained on our own artificial data set which imitates images from the Fermi Large Area Telescope. These images are raw count photon maps of 10x10 degrees covering energies from 1 to 10 GeV. We compare different CNN architectures that demonstrate accuracy increase by ~15% and reduces the inference time by at least the factor of 4 accuracy improvement with respect to a similar state of the art models.