[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85826-en":3,"doc-seo-85826-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},85826,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","FlowPainter: Inpainting Optical Flow via Confidence-Guided Completion","Optical flow estimation aims to infer dense pixel-wise motion between consecutive frames, yet remains difficult under fast motion, occlusions, illumination changes, and textureless regions. Existing approaches split into iterative refinement and diffusion-based generation, but diffusion models denoise the entire flow field even in regions where motion structure is already reliably estimable, increasing denoising cost and risking slow convergence. FlowPainter reformulates dense-flow generation as confidence-guided soft inpainting using a confidence mask, confidence-based initialization, and confidence-gated residual guidance to stabilize early denoising while refining late details.","arXiv :2607 . 10140v1 [ cs .CV] 11 Jul 2026  \nFlowPainter: Inpainting Optical Flow via Confidence-Guided Completion  \nYuang Meng 1 ,3 , Chenyang Wu 1 , Xianshun Liu 1 ,3 , Chun-Le Guo 1 ⋆ , Zichen Liang 1 , Lina Lei 1 , Jie Liang3 , Hui Zeng3 , Chongyi Li 1 , and Lei Zhang2 ,3  \n1 VCIP, CS, Nankai University  \n2 The Hong Kong Polytechnic University  \n3 OPPO Research Institute  \n[mya@mail.nankai.edu.cn](mya@mail.nankai.edu.cn)  \nAbstract. Existing optical flow estimation methods broadly follow two paradigms: iterative optimization and diffusion-based estimation. Iterative methods, exemplified by RAFT, achieve accurate flow estimation through recurrent refinement, but can still be challenged by large displacements and complex motion patterns. Diffusion-based methods introduce generative modeling into optical flow and have shown promising results in such ambiguous regions. However, existing diffusion-based flow models usually denoise the entire dense flow field from Gaussian noise, including simple regions where reliable motion structure can already be estimated by a lightweight network. This increases the denoising burden and may lead to slow convergence and unstable training. To address these issues, we introduce FlowPainter, a diffusion-based optical flow framework that reformulates dense-flow generation as confidence-guided soft inpainting.  \nFlowPainter first employs a lightweight confidence-aware network to predict a rough flow and a pixel-wise confidence mask, which serves as a reliability gate for distinguishing reliable simple regions from uncertain hard regions. The resulting simple-flow prior is used for confidence-based initialization and is further injected into the iterative denoising process through confidence-gated residual guidance. With a dynamically decaying guidance strength, FlowPainter stabilizes early denoising while preserving the flexibility of the diffusion model for late-stage detail refinement. Extensive experiments on public benchmarks, including Sintel, KITTI, and Spring, demonstrate that FlowPainter achieves strong accuracy under comparable training settings and improves convergence efficiency over existing diffusion-based optical flow methods, with notable gains on challenging benchmark splits. Our approach provides a practical direction for integrating reliable discriminative priors with diffusion-based refinement for optical flow estimation. Our code is made publicly available at [https://github.com/mya012/FlowPainter](https://github.com/mya012/FlowPainter).  \nKeywords: Optical flow · Diffusion model · Inpainting  \n⋆ C. L. Guo is the corresponding author.  \n2 Y. Meng et al.  \n1 Introduction  \nOptical flow estimation is a fundamental problem in computer vision, aiming to infer a dense motion vector for each pixel between consecutive frames. It is widely used in video understanding [8,16], action recognition [26,36], autonomous driving [2, 19, 32], video frame interpolation [10, 29, 42], and restoration problems [3, 22, 28] . Despite decades of progress, accurate optical flow estimation remains challenging in real-world scenarios due to fast motion, occlusions, illumination changes, and textureless regions.  \nDeep learning has greatly advanced optical flow estimation. Early CNNbased methods [7,13,30,35] introduced end-to-end regression, stacked refinement, and coarse-to-fine matching, but their fixed receptive fields and limited feature expressiveness can hinder complex motion modeling. RAFT [37] established a strong iterative refinement paradigm by building an all-pairs correlation volume and recurrently updating the flow field. Its follow-up works [15, 27, 38, 41] further improve global aggregation, 3D motion modeling, training strategies, and practical efficiency. Nevertheless, iterative optimization can still be challenged by very large displacements, severe occlusions, and complex non-rigid motion, where correlation construction and recurrent updates may become less reliable [17] .  \nIn parall","cbCairYwu35oyUD1","https://ap.wps.com/l/cbCairYwu35oyUD1","pdf",2855007,1,18,"English","en",105,"# Introduction\n## Optical flow estimation challenges\n## Iterative optimization paradigm\n## Diffusion-based flow estimation","[{\"question\":\"What limitation do existing diffusion-based optical flow models face?\",\"answer\":\"They typically denoise the entire dense flow field from Gaussian noise, including simple regions where reliable motion can already be estimated by a lightweight network. This increases denoising burden and can slow convergence or destabilize training.\"},{\"question\":\"How does FlowPainter reformulate optical flow generation?\",\"answer\":\"FlowPainter reformulates dense-flow generation as confidence-guided soft inpainting. It uses a confidence-aware network to predict a rough flow and a pixel-wise confidence mask that separates reliable simple regions from uncertain hard regions.\"},{\"question\":\"How does FlowPainter use confidence during the denoising process?\",\"answer\":\"It initializes the process with a simple-flow prior derived from confidence-based outputs and injects this prior into iterative denoising via confidence-gated residual guidance. The guidance strength dynamically decays to stabilize early denoising and preserve late-stage detail refinement flexibility.\"}]",1784206503,45,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"flowpainter-inpainting-optical-flow-via-confidence-guided-completion","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/flowpainter-inpainting-optical-flow-via-confidence-guided-completion/85826/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What limitation do existing diffusion-based optical flow models face?","Question",{"text":75,"@type":76},"They typically denoise the entire dense flow field from Gaussian noise, including simple regions where reliable motion can already be estimated by a lightweight network. This increases denoising burden and can slow convergence or destabilize training.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does FlowPainter reformulate optical flow generation?",{"text":80,"@type":76},"FlowPainter reformulates dense-flow generation as confidence-guided soft inpainting. It uses a confidence-aware network to predict a rough flow and a pixel-wise confidence mask that separates reliable simple regions from uncertain hard regions.",{"name":82,"@type":73,"acceptedAnswer":83},"How does FlowPainter use confidence during the denoising process?",{"text":84,"@type":76},"It initializes the process with a simple-flow prior derived from confidence-based outputs and injects this prior into iterative denoising via confidence-gated residual guidance. The guidance strength dynamically decays to stabilize early denoising and preserve late-stage detail refinement flexibility.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]