Lynge Lauritsen
Gather.town id
MLA11
Poster Title
Super-resolving Herschel SPIRE images using Convolutional Neural Networks
Institution
The Open University
Abstract (short summary)
Wide-field sub-millimetre surveys have driven many major advances in galaxy evolution in the past decade, but without extensive follow-ups the coarse angular resolution of these surveys limits the science exploitation. This has driven many deconvolution efforts. Generative Adversarial Networks have already been used to attempt deconvolutions on optical data. In this talk I will present an autoencoder with a novel loss function to overcome this problem at submm wavelengths. This approach is successfully demonstrated on Herschel SPIRE COSMOS data, with the super-resolving target being the JCMT SCUBA-2 observations of the same field. We reproduce the JCMT SCUBA-2 images with surprisingly high fidelity, and quantify the point source flux constraints using this autoencoder.
Plain text (extended) Summary
Wide-field sub-millimetre surveys have driven many major advances in galaxy evolution in the past decade, but without extensive follow-up observations the coarse angular resolution of these surveys limits the science exploitation. This has driven the development of various analytical deconvolution methods. Here we present an autoencoder with a novel loss function to overcome this problem in the sub-millimeter wavelength range. This approach is successfully demonstrated on Herschel SPIRE 500 µm COSMOS data, with the super-resolving target being the JCMT SCUBA-2 450 µm observations of the same field. We reproduce the JCMT SCUBA-2 images with high fidelity using this autoencoder. This is quantified through the point source fluxes and positions, the completeness and the purity.
URL
lynge.lauritsen@open.ac.uk
Poster file