{"paper":{"title":"SyncDreamer: Generating Multiview-consistent Images from a Single-view Image","license":"http://creativecommons.org/licenses/by/4.0/","headline":"SyncDreamer generates multiple consistent views of an object from one input image by synchronizing their diffusion process.","cross_cats":["cs.AI","cs.GR"],"primary_cat":"cs.CV","authors_text":"Cheng Lin, Lingjie Liu, Taku Komura, Wenping Wang, Xiaoxiao Long, Yuan Liu, Zijiao Zeng","submitted_at":"2023-09-07T02:28:04Z","abstract_excerpt":"In this paper, we present a novel diffusion model called that generates multiview-consistent images from a single-view image. Using pretrained large-scale 2D diffusion models, recent work Zero123 demonstrates the ability to generate plausible novel views from a single-view image of an object. However, maintaining consistency in geometry and colors for the generated images remains a challenge. To address this issue, we propose a synchronized multiview diffusion model that models the joint probability distribution of multiview images, enabling the generation of multiview-consistent images in a s"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"SyncDreamer generates images with high consistency across different views, thus making it well-suited for various 3D generation tasks such as novel-view-synthesis, text-to-3D, and image-to-3D.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The 3D-aware feature attention mechanism successfully correlates corresponding features across views without introducing new inconsistencies or artifacts during the joint reverse diffusion process.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SyncDreamer produces multiview-consistent images from a single input image by jointly modeling their distribution and synchronizing intermediate diffusion states via 3D-aware attention.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"SyncDreamer generates multiple consistent views of an object from one input image by synchronizing their diffusion process.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"6a71eaed771d41734b94558cb31517e09f0c28dcf5415fa48552c39ab7a00e44"},"source":{"id":"2309.03453","kind":"arxiv","version":2},"verdict":{"id":"284650cd-755d-428e-9662-45d10de30678","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T10:31:44.910240Z","strongest_claim":"SyncDreamer generates images with high consistency across different views, thus making it well-suited for various 3D generation tasks such as novel-view-synthesis, text-to-3D, and image-to-3D.","one_line_summary":"SyncDreamer produces multiview-consistent images from a single input image by jointly modeling their distribution and synchronizing intermediate diffusion states via 3D-aware attention.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The 3D-aware feature attention mechanism successfully correlates corresponding features across views without introducing new inconsistencies or artifacts during the joint reverse diffusion process.","pith_extraction_headline":"SyncDreamer generates multiple consistent views of an object from one input image by synchronizing their diffusion process."},"references":{"count":46,"sample":[{"doi":"","year":null,"title":"Re-imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond","work_id":"1040fcd2-7a4c-4a7a-8803-bad82a00eea4","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Multidiffusion: Fusing diffusion paths for controlled image generation","work_id":"1a06e9b7-e97f-4665-ba6e-55301822d3b6","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"ShapeNet: An Information-Rich 3D Model Repository","work_id":"b2ac5b60-daa9-435b-9369-12271e126edd","ref_index":3,"cited_arxiv_id":"1512.03012","is_internal_anchor":true},{"doi":"","year":null,"title":"Single- stage diffusion nerf: A unified approach to 3d generation and reconstruction","work_id":"f4a4c59d-3c8e-4a0d-97cd-a540bc93203f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Objaverse-XL: A Universe of 10M+ 3D Objects","work_id":"1c5475ad-d1ec-4de1-8670-b8cd5a4c85d3","ref_index":5,"cited_arxiv_id":"2307.05663","is_internal_anchor":true}],"resolved_work":46,"snapshot_sha256":"55f35f9925fd55672eba4a2833bff437bffffbb1e69aed2944add4e63b8d372f","internal_anchors":7},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c16a268164c24fa21927a81e9233b242eabba7083285aaed2820eeed18a8ee81"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}