HELP: XID+, The Probabilistic De-blender for Herschel SPIRE maps
P.D. Hurley, S.Oliver, M.Betancourt, C.Clarke, W.I.Cowley,, S.Duivenvoorden, D.Farrah, M.Griffin, C.Lacey, E.Le Floc'h, A.Papadopoulos,, M.Sargent, J.M.Scudder, M.Vaccari, I. Valtchanov, L.Wang

TL;DR
XID+ is a Bayesian source extraction tool for Herschel SPIRE maps that improves flux estimation accuracy by incorporating prior information and providing full posterior distributions, outperforming previous methods.
Contribution
The paper introduces XID+, a novel probabilistic Bayesian tool for Herschel SPIRE source extraction that offers improved flux accuracy and uncertainty estimation over existing methods.
Findings
XID+ outperforms DESPHOT in flux accuracy.
It provides full posterior distributions for flux estimates.
Application to COSMOS maps demonstrates its effectiveness.
Abstract
We have developed a new prior-based source extraction tool, XID+, to carry out photometry in the Herschel SPIRE maps at the positions of known sources. XID+ is developed using a probabilistic Bayesian framework which provides a natural framework in which to include prior information, and uses the Bayesian inference tool, Stan, to obtain the full posterior probability distribution on flux estimates. In this paper, we discuss the details of XID+ and demonstrate the basic capabilities and performance by running it on simulated SPIRE maps resembling the COSMOS field, and comparing to the current prior-based source extraction tool DESPHOT. We show that not only does XID+ perform better on metrics such as flux accuracy and flux uncertainty accuracy, we illustrate how obtaining the posterior probability distribution can help overcome some of the issues inherent with maximum likelihood based…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
