{"paper":{"title":"A Survey on Diffusion Models for Inverse Problems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Pre-trained diffusion models serve as unsupervised priors to solve inverse problems such as image restoration and reconstruction without any additional training.","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Alexandros G. Dimakis, Chieh-Hsin Lai, Giannis Daras, Hyungjin Chung, Jong Chul Ye, Mauricio Delbracio, Peyman Milanfar, Yuki Mitsufuji","submitted_at":"2024-09-30T17:34:01Z","abstract_excerpt":"Diffusion models have become increasingly popular for generative modeling due to their ability to generate high-quality samples. This has unlocked exciting new possibilities for solving inverse problems, especially in image restoration and reconstruction, by treating diffusion models as unsupervised priors. This survey provides a comprehensive overview of methods that utilize pre-trained diffusion models to solve inverse problems without requiring further training. We introduce taxonomies to categorize these methods based on both the problems they address and the techniques they employ. We ana"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"This survey provides a comprehensive overview of methods that utilize pre-trained diffusion models to solve inverse problems without requiring further training. We introduce taxonomies to categorize these methods based on both the problems they address and the techniques they employ.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the selected methods and introduced taxonomies accurately and comprehensively represent the current landscape of diffusion-based approaches to inverse problems without major omissions or biases in literature coverage.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A survey that introduces taxonomies for categorizing pre-trained diffusion model methods applied to inverse problems and analyzes their connections and challenges.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Pre-trained diffusion models serve as unsupervised priors to solve inverse problems such as image restoration and reconstruction without any additional training.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4b85f29189d879f79b99b6a4c2424e0936839d5c541eb8d62659d41be290bc3d"},"source":{"id":"2410.00083","kind":"arxiv","version":1},"verdict":{"id":"3ee9cea5-94d3-4aa5-aee7-dc775a4776e3","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T04:27:39.886408Z","strongest_claim":"This survey provides a comprehensive overview of methods that utilize pre-trained diffusion models to solve inverse problems without requiring further training. We introduce taxonomies to categorize these methods based on both the problems they address and the techniques they employ.","one_line_summary":"A survey that introduces taxonomies for categorizing pre-trained diffusion model methods applied to inverse problems and analyzes their connections and challenges.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the selected methods and introduced taxonomies accurately and comprehensively represent the current landscape of diffusion-based approaches to inverse problems without major omissions or biases in literature coverage.","pith_extraction_headline":"Pre-trained diffusion models serve as unsupervised priors to solve inverse problems such as image restoration and reconstruction without any additional training."},"references":{"count":165,"sample":[{"doi":"","year":2021,"title":"Robust compressed sensing mri with deep generative priors,","work_id":"4506462d-8ec0-4a29-becd-b3ff204fb442","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2011,"title":"Score-Based Generative Modeling through Stochastic Differential Equations","work_id":"d9110e53-a5d4-4794-a4c5-a575e91c31ad","ref_index":2,"cited_arxiv_id":"2011.13456","is_internal_anchor":true},{"doi":"","year":2021,"title":"Ilvr: Cond itioning method for denoising diffusion probabilistic models,","work_id":"5a836a66-4111-48fa-a13a-6f2952d784a0","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"D iffu- sion posterior sampling for general noisy inverse problems ,","work_id":"d831b21d-35fa-497e-8491-7ed0c80e572f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Pseudoinve rse-guided diffusion models for inverse problems,","work_id":"272012b4-e027-45b4-8955-0898d1a0ff8c","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":165,"snapshot_sha256":"5037af0d66a61393f39cbd109aafcb6bd63f72634f2471dae73313b83e159702","internal_anchors":6},"formal_canon":{"evidence_count":2,"snapshot_sha256":"29ca2ac88be215484dd4f4c23551292700e27938b194ecd4ea5bd0c7f65d9e50"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}