Score-matching neural networks for improved multi-band source separation
arXiv:2401.07313 · doi:10.1016/j.ascom.2024.100875
Abstract
We present the implementation of a score-matching neural network that represents a data-driven prior for non-parametric galaxy morphologies. The gradients of this prior can be incorporated in the optimization of galaxy models to aid with tasks like deconvolution, inpainting or source separation. We demonstrate this approach with modification of the multi-band modeling framework Scarlet that is currently employed as deblending method in the pipelines of the HyperSuprimeCam survey and the Rubin Observatory. The addition of the prior avoids the requirement of non-differentiable constraints, which can lead to convergence failures we discovered in Scarlet. We present the architecture and training details of our score-matching neural network and show with simulated Rubin-like observations that using a data-driven prior outperforms the baseline \scarlet method in accuracy of total flux and morphology estimates, while maintaining excellent performance for colors. We also demonstrate significant improvements in the robustness to inaccurate initializations. The trained score models used for this analysis are publicly available at https://github.com/SampsonML/galaxygrad.
17 pages, 13 figures, accepted in Astronomy and Computing
References in corpus (23)
- Denoising Diffusion Probabilistic Models
- LSST: from Science Drivers to Reference Design and Anticipated Data Products
- The Zwicky Transient Facility: System Overview, Performance, and First Results
- Score-Based Generative Modeling through Stochastic Differential Equations
- Generative Modeling by Estimating Gradients of the Data Distribution
- The Zwicky Transient Facility: Science Objectives
- Weak lensing for precision cosmology
- The Zwicky Transient Facility: Observing System
- A Proximal Decomposition Method for Solving Convex Variational Inverse Problems
- DES Y3 results: Blending shear and redshift biases in image simulations
- Proximal Splitting Methods in Signal Processing
- Deblending and Classifying Astronomical Sources with Mask R-CNN Deep Learning
- SCARLET: Source separation in multi-band images by Constrained Matrix Factorization
- Photometry of high-redshift blended galaxies using deep learning
- Deblending galaxy superpositions with branched generative adversarial networks
- Deblending galaxies with Variational Autoencoders: a joint multi-band, multi-instrument approach
- Multi-band morpho-Spectral Component Analysis Deblending Tool (MuSCADeT): Deblending colourful objects
- Deblending Galaxies with Generative Adversarial Networks
- Galaxy Deblending using Residual Dense Neural networks
- Proximal Adam: Robust Adaptive Update Scheme for Constrained Optimization
- Learning Diffusion Priors from Observations by Expectation Maximization
- Spotting Hallucinations in Inverse Problems with Data-Driven Priors
- Panic at the ISCO: time-varying double-peaked broad lines from evolving accretion disks are common amongst optically variable AGN