[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82184-en":3,"doc-seo-82184-105":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},82184,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","REBASE Reference Background Subspace Elimination for Training Free In-Context Segmentation","Training-free in-context segmentation lets new object categories be introduced at inference using a single annotated reference image, avoiding retraining and class-incremental memory overhead. Existing methods typically combine semantic correspondence from vision foundation models with promptable segmentation networks such as SAM, but are limited by cross-image similarity maps: shared contextual backgrounds create spurious similarity in non-target regions, reducing prompt localization quality. REBASE suppresses these effects by removing low-rank background feature subspaces in closed form, improving semantic matching and one-shot masks without training.","REBASE: Reference-Background Subspace Elimination for Training-Free  \nIn-Context Segmentation  \nMantha Sai Gopal Jaison Saji Chacko Harsh Nandwana Sandesh Hegde Debarshi Banerjee Uma Mahesh  \nCamCom Technologies Private Limited*  \narXiv :2607 .09082v1 [ cs .CV] 10 Jul 2026  \nAbstract  \nTraining-free in-context segmentation enables new object categories to be introduced at inference time from a single annotated reference image, eliminating the retraining and memory overhead of class-incremental learning. Recent approaches achieve this by combining vision foundation models for semantic correspondence with promptable segmentation networks like SAM. However, their performance is fundamentally limited by the quality of the cross-image similarity map; shared contextual backgrounds between the reference and query systematically elevate similarity in non-target regions, degrading prompt localization. We present REBASE, a training-free framework that explicitly suppresses these spurious contextual correspondences. Our method identifies the low-rank background feature subspace from the reference image and project the reference and query features onto its orthogonal complement in closed form, yielding cleaner semantic matching. We then generate positive point prompts using similarity-weighted farthest-point sampling, paired with a refined dense similarity prior. Without any training or parameter updates, our approach establishes a new state of theart among training-free methods on PACO-Part, FSS-1000, and cross-domain datasets such as ISIC2018, demonstrating that explicit background subspace removal is a highly effective principle for one-shot localization.Code is released at:  \n[https://github.com/ai-and-lab/rebase](https://github.com/ai-and-lab/rebase)  \n1. Introduction  \nReal-world deployment of a vision system rarely involves a fixed, closed set of object categories. New instances of interest such as a customer’s specific product, a rare medical structure or a domain-specific part arrive continuously after the model has been deployed. Class-incremental learning  \n*Emails: {saigopal .mantha, jaison .saji, harsh.nandwana, sandesh.hegde,  \ndebarshi.banerjee, [umesh](umesh}@camcom.ai)[}](umesh}@camcom.ai)[@camcom.ai](umesh}@camcom.ai)  \nFigure 1 . Qualitative examples of our training-free one-shot segmentation method, demonstrating cleaner object boundaries than GF-SAM and INSID3 across two segmentation tasks (top and bottom rows) . Each row shows, from left to right, the reference image with its annotation, the query image with its ground truth, and the predictions produced by GF-SAM, INSID3, and Ours.  \n(CIL) is the canonical framework for this setting, wherein a recognition or segmentation model is repeatedly fine-tuned as new classes appear, with mechanisms to mitigate catastrophic forgetting [12, 30] . However, in CIL, each new class incurs an optimization pass over potentially the full training set, the storage of either replay data or parameter snapshots, and a non-trivial hyperparameter budget per increment. Moreover, the system is unable to serve a new class until the next fine-tuning step has completed.  \nFew-shot segmentation on the other hand, offers a structurally different solution. The target class is specified by a small set of annotated reference examples at inference time, and the model is tasked with segmenting the target in the query images without requiring any further training or parameter updates [7, 19, 23] . In the one-shot (1-shot) regime, in particular, a single reference image with a binary mask must suffice. This setting subsumes the practical desiderata of CIL. New classes can be added on demand, with zero re-training and zero cross-class interference.  \nThe practical realization of this training-free paradigm has been enabled by recent advances in vision foundation models, which naturally decompose the problem into semantic  \ncorrespondence and geometric localization. Self-supervised vision transformers, su","cbCaikUNKqOT8mAh","https://ap.wps.com/l/cbCaikUNKqOT8mAh","pdf",17556174,1,16,"English","en",105,"# Introduction\n## Training-free in-context segmentation problem\n## Limitations of similarity maps and contextual background\n## Proposed REBASE approach","[{\"question\":\"What problem does REBASE address in training-free in-context segmentation?\",\"answer\":\"REBASE addresses the degradation caused by spurious contextual correspondences in cross-image similarity maps, where shared backgrounds inflate similarity outside the target region and worsen prompt localization.\"},{\"question\":\"How does REBASE improve semantic correspondence without training updates?\",\"answer\":\"REBASE finds the low-rank background feature subspace from the reference image and projects reference and query features onto its orthogonal complement in closed form, producing cleaner semantic matching.\"},{\"question\":\"What is the key idea behind generating prompts in REBASE?\",\"answer\":\"REBASE generates positive point prompts using similarity-weighted farthest-point sampling and uses a refined dense similarity prior, enabling accurate one-shot mask localization without parameter updates.\"}]",1784178666,40,{"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},"rebase-reference-background-subspace-elimination-for-training-free-in-context-segmentation","",{"@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/rebase-reference-background-subspace-elimination-for-training-free-in-context-segmentation/82184/",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 problem does REBASE address in training-free in-context segmentation?","Question",{"text":75,"@type":76},"REBASE addresses the degradation caused by spurious contextual correspondences in cross-image similarity maps, where shared backgrounds inflate similarity outside the target region and worsen prompt localization.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does REBASE improve semantic correspondence without training updates?",{"text":80,"@type":76},"REBASE finds the low-rank background feature subspace from the reference image and projects reference and query features onto its orthogonal complement in closed form, producing cleaner semantic matching.",{"name":82,"@type":73,"acceptedAnswer":83},"What is the key idea behind generating prompts in REBASE?",{"text":84,"@type":76},"REBASE generates positive point prompts using similarity-weighted farthest-point sampling and uses a refined dense similarity prior, enabling accurate one-shot mask localization without parameter updates.","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},[93,97,101,105,110,115,119,122,127,130,134],{"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":28,"slug":118},7,"Healthcare","healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]