{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deblending-galaxy-superpositions-with","title":"Deblending galaxy superpositions with branched generative adversarial networks","arxiv_id":"1810.10098","date":"2018-10-23","proceeding":null,"authors":["David M. Reiman","Brett E. Göhre"],"abstract":"Near-future large galaxy surveys will encounter blended galaxy images at a\nfraction of up to 50% in the densest regions of the universe. Current\ndeblending techniques may segment the foreground galaxy while leaving missing\npixel intensities in the background galaxy flux. The problem is compounded by\nthe diffuse nature of galaxies in their outer regions, making segmentation\nsignificantly more difficult than in traditional object segmentation\napplications. We propose a novel branched generative adversarial network (GAN)\nto deblend overlapping galaxies, where the two branches produce images of the\ntwo deblended galaxies. We show that generative models are a powerful engine\nfor deblending given their innate ability to infill missing pixel values\noccluded by the superposition. We maintain high peak signal-to-noise ratio and\nstructural similarity scores with respect to ground truth images upon\ndeblending. Our model also predicts near-instantaneously, making it a natural\nchoice for the immense quantities of data soon to be created by large surveys\nsuch as LSST, Euclid and WFIRST.","url_abs":"http://arxiv.org/abs/1810.10098v4","url_pdf":"http://arxiv.org/pdf/1810.10098v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deblending-galaxy-superpositions-with","repo_url":"https://github.com/davidreiman/deblender","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":null,"task_name":"Generative Adversarial Network"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}