Waste-free sequential Monte Carlo for light source detection in astronomical images

Waste-free sequential Monte Carlo (SMC) is an extension of SMC that was recently proposed by Dau and Chopin (2022). In this report, we conduct numerical experiments to assess the utility of embedding this waste-free procedure within an SMC sampler designed for a challenging Bayesian inverse problem that requires transdimensional inference: detecting and distinguishing overlapping light sources in astronomical images.