[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-346647-105":59,"doc-detail-346647-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","you-only-erase-once-object-erasure-diffusion-model","You Only Erase Once - Object Erasure Diffusion Model","","YOEO presents an object erasure method designed to remove target objects without introducing unwanted objects or artifacts in masked regions. Unlike recent diffusion-based approaches that rely on limited paired data and explicit constraints, YOEO trains an object erasure diffusion model using unpaired large-scale real-world images. A sundries detector and a context coherence loss, built on entity segmentation and feature coherence from a pretrained segmentation network, provide pair-free supervision. A diffusion distillation strategy enables efficient few-step inference while achieving state-of-the-art erasure quality.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/you-only-erase-once-object-erasure-diffusion-model/346647/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/you-only-erase-once-object-erasure-diffusion-model/346647.png","ImageObject",300,407,{"name":92,"@type":93},"Mary Man","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-29","2026-09-22",true,{"@type":102,"interactionType":103,"userInteractionCount":52},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What is the main goal of YOEO in object erasure?","Question",{"text":112,"@type":113},"YOEO removes selected objects in a single pass while preventing unexpected objects or artifacts from appearing inside the masked regions, and while preserving contextual coherence with surrounding content.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"Why do diffusion-based erasure methods often fail in practice?",{"text":117,"@type":113},"They can hallucinate unintended objects after removal and often rely on synthetic paired data or supervised fine-tuning that mainly teaches denoising rather than the task-specific erasure objective with contextual consistency.",{"name":119,"@type":110,"acceptedAnswer":120},"How does YOEO provide supervision without paired erased ground truth?",{"text":121,"@type":113},"YOEO uses pair-free erasure-related supervision: a sundries detector evaluates unwanted artifacts, and a context coherence loss measures feature coherence between completed and original regions using features from a pretrained segmentation network.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},346647,1790516747,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":52,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145,"preview_page_count":19,"preview_urls":146},7421720224475,"https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c","This CVPR paper is the Open Access version, provided by the Computer Vision Foundation.  \nExcept for this watermark, it is identical to the accepted version; the final published version of the proceedings is available on IEEE Xplore.  \nYou Only Erase Once: Erasing Anything without Bringing Unexpected Content  \nYixing Zhu 1 Qing Zhang 1 ,3∗ Wenju Xu2 Wei-Shi Zheng 1 ,3  \n1 School of Computer Science and Engineering, Sun Yat-sen University, China 2Amazon  \n3 Key Laboratory of Machine Intelligence and Advanced Computing, Ministry of Education, China  \nInpainting mask  \nFigure 1 . Comparison of object erasure. Unlike the compared methods which tend to generate unwanted objects or artifacts within themasked regions, our method allows to cleanly erase the target object in a single pass, without introducing any unexpected content while maintaining the overall harmony and context consistency.  \nAbstract  \nWe present YOEO, an approach for object erasure. Unlike recent diffusion-based methods which struggle to erase target objects without generating unexpected content within the masked regions due to lack of sufficient paired training data and explicit constraint on content generation, our method allows to produce high-quality object erasure results free of unwanted objects or artifacts while faithfully preserving the overall context coherence to the surrounding content. We achieve this goal by training an object erasure diffusion model on unpaired data containing only  \n∗ Corresponding author ([zhangq93@mail.sysu.edu.cn](zhangq93@mail.sysu.edu.cn)).  \nlarge-scale real-world images, under the supervision of a sundries detector and a context coherence loss that are built upon an entity segmentation model. To enable more efficient training and inference, a diffusion distillation strategy is employed to train for a few-step erasure diffusion model. Extensive experiments show that our method outperforms the state-of-the-art object erasure methods. Our code is available at [https://zyxunh.github.io/YOEO-ProjectPage/](https://zyxunh.github.io/YOEO-ProjectPage/).  \n1. Introduction  \nRecent advances in diffusion models [5, 8, 14, 18, 30, 33] have greatly advanced image inpainting, enabling applica-  \ntions such as restoration, completion, and editing. Among these, object erasure has become an essential capability for visual editing and content refinement. Previously, most erasure methods relied on GAN-based frameworks [12], which often produced blurry textures and artifacts. In contrast, diffusion-based approaches can generate visually realistic results with rich details and textures, and have therefore become the mainstream solution.  \nHowever, diffusion-based methods often hallucinate by inserting unintended objects after removing the target ones, leading to contextually inconsistent results (see Figure 1) . On the other hand, recent closed-source multimodal models such as ChatGPT and Nano Banana, though are more powerful in object erasure, but involve large parameter countsand high computational overhead, hindering their practical deployment on edge devices. Hence, it is quite necessary to develop a dedicated object erasure model that not only enables superior erasure performance but also enjoys low inference latency and significantly fewer parameters.  \nSeveral specialized erasure methods try to mitigate hallucination issues by modifying attention probability maps [20, 23], generating synthetic data [15], introducing segmentation [57] or geometric [58] priors. However, these methods still struggle in complex open-world scenes, often producing incoherent or unrealistic content. We identify two main causes of this limitation. First, most diffusion-based erasure models are trained exclusively on synthetic paired data created by partially masking or inserting objects into clean images and using the original image as the ground truth. Such data fail to represent real-world erasure scenarios, where ground-truth erased results are unavailable. Sec","cbCaisEURidIVK0j","https://ap.wps.com/l/cbCaisEURidIVK0j","pdf",12213365,11,"English","# Abstract\n# 1. Introduction\n## Problem of hallucinated content in diffusion-based erasure\n## Limitations of existing training and supervision\n## Proposed framework: You Only Erase Once (YOEO)\n## Key idea: pair-free erasure supervision with coherence constraints\n## Distillation for efficient few-step denoising","[{\"question\":\"What is the main goal of YOEO in object erasure?\",\"answer\":\"YOEO removes selected objects in a single pass while preventing unexpected objects or artifacts from appearing inside the masked regions, and while preserving contextual coherence with surrounding content.\"},{\"question\":\"Why do diffusion-based erasure methods often fail in practice?\",\"answer\":\"They can hallucinate unintended objects after removal and often rely on synthetic paired data or supervised fine-tuning that mainly teaches denoising rather than the task-specific erasure objective with contextual consistency.\"},{\"question\":\"How does YOEO provide supervision without paired erased ground truth?\",\"answer\":\"YOEO uses pair-free erasure-related supervision: a sundries detector evaluates unwanted artifacts, and a context coherence loss measures feature coherence between completed and original regions using features from a pretrained segmentation network.\"}]","You Only Erase Once - Object Erasure Diffusion Model | PDF",1790062277,28,[147,151,155,159],{"page_no":8,"url":148,"width":149,"height":150},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/ff76b729ac055d08e290377ba6ce8e6c.webp",983,1272,{"page_no":14,"url":152,"width":153,"height":154},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/76aecc5b8c45564764e69ed52a394052.webp",982,1270,{"page_no":81,"url":156,"width":157,"height":158},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/3af251bb3ace35c75c5dc977c9ec5d73.webp",981,1269,{"page_no":19,"url":160,"width":161,"height":162},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/136da015afaeadad258be1aff9641026.webp",980,1268]