{"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/rare-image-reconstruction-using-deep-priors","title":"RARE: Image Reconstruction using Deep Priors Learned without Ground Truth","arxiv_id":"1912.05854","date":"2020-05-23","proceeding":null,"authors":[],"abstract":"Regularization by denoising (RED) is an image reconstruction framework that\nuses an image denoiser as a prior. Recent work has shown the state-of-the-art\nperformance of RED with learned denoisers corresponding to pre-trained\nconvolutional neural nets (CNNs). In this work, we propose to broaden the\ncurrent denoiser-centric view of RED by considering priors corresponding to\nnetworks trained for more general artifact-removal. The key benefit of the\nproposed family of algorithms, called regularization by artifact-removal\n(RARE), is that it can leverage priors learned on datasets containing only\nundersampled measurements. This makes RARE applicable to problems where it is\npractically impossible to have fully-sampled groundtruth data for training. We\nvalidate RARE on both simulated and experimentally collected data by\nreconstructing a free-breathing whole-body 3D MRIs into ten respiratory phases\nfrom heavily undersampled k-space measurements. Our results corroborate the\npotential of learning regularizers for iterative inversion directly on\nundersampled and noisy measurements.","url_abs":"http://arxiv.org/abs/1912.05854v2","url_pdf":"http://arxiv.org/pdf/1912.05854v2.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":"rare-image-reconstruction-using-deep-priors","repo_url":"https://github.com/wustl-cig/RARE","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"image-reconstruction","task_name":"Image Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1912.05854","atlas_url":"https://app.syntology.ai/?focus=1912.05854","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}