Equivariant Neural Networks for Dark Matter Morphology with Strong Gravitational Lensing

The study of substructures in the dark matter has shown signs of promise to deliver on the open-ended and long-standing problem of the identity of dark matter. To probe these substructures, strong gravitational lensing has been used in the past and provided some interesting results. The approaches based on deep learning have the ability to identify and differentiate among these substructures using images from the simulation of strong gravitational lensing. This project explores the scope of use of equivariant neural networks that can benefit from inherent symmetry present in natural images. We will be implementing and benchmarking the results of equivariant neural networks on the available DeepLens simulated dataset. We will also integrate all these architectures with the DeepLens pipeline to provide a high-level interface for future work.