[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81524-en":3,"doc-seo-81524-105":29,"detail-sidebar-cat-0-en-105":90},{"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":4,"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},81524,1099513958762,"Logic","https://ap-avatar.wpscdn.com/avatar/1000023916a998db790?x-image-process=image/resize,m_fixed,w_180,h_180&k=1782109480056885918",8,"Research & Report","AffordanceSAM: Segment Anything Once More in Affordance Grounding","Affordance grounding aims to localize actionable “action possibilities” on objects to bridge visual perception and robotic action. The document identifies limitations in weakly supervised approaches (complex training and lack of new action inference) and in fully supervised ones (poor generalization under limited annotations and components trained from scratch). It proposes AffordanceSAM, extending SAM-style segmentation generalization to affordance grounding via an affordance-adaption module and a coarse-to-fine dataset, C2F-Aff, trained in three stages. Results confirm state-of-the-art performance on AGD20K and strong transferability.","AffordanceSAM: Segment Anything Once More in Affordance Grounding  \nDengyang Jiang2 ,3 , 1∗† Zanyi Wang4 , 1∗† Hengzhuang Li5 , 1∗ Sizhe Dang 1 Teli Ma3 Wei Wei2 Guang Dai 1 Lei Zhang2 Harry Yang3 Mengmeng Wang6 , 1 ✉  \n1 SGIT AI Lab 2NWPU 3HKUST 4XJTU 5HUST 6 ZJUT  \narXiv :2504 . 15650v3 [ cs .CV] 10 Jul 2026  \nAbstract  \nBuilding a generalized affordance grounding model to identify actionable regions on objects is vital for real-world applications. Existing methods to train the model can be divided into weakly and fully supervised ways. However, the former method requires a complex training framework design and can not infer new actions without an auxiliary prior. While the latter often struggle with limited annotated data and components trained from scratch despite being simpler. This study focuses on fully supervised affordance grounding and overcomes its limitations by proposing AffordanceSAM, which extends SAM’s generalization capacity in segmentation to affordance grounding. Specifically, we design an affordanceadaption module and curate a coarse-to-fine annotated dataset called C2F-Aff to thoroughly transfer SAM’s robust performance to affordance in a three-stage training manner. Experimental results confirm that AffordanceSAM achieves stateof-the-art (SOTA) performance on the AGD20K benchmark and exhibits strong generalized capacity.  \nIntroduction  \nAffordance grounding (Gibson 2014) refers to finding potential “action possibilities” regions of an object, which plays a key role in bridging the gap between visual perception and robotic action. Recently, attempts made to endow models to have such grounding abilities can be broadly divided into two ways. The first type of methods is weakly supervised affordance grounding (Li et al. 2023a; Xu and Mu 2025; Wang et al. 2025), which aims to identify affordance regions on objects using human-object interaction images and egocentric object images without dense labels. The affordance maps are obtained via class activation mapping (CAM) (Zhou et al. 2016) or an auxiliary prior model (Kirillov et al. 2023) . However, these methods require a complex training framework design since there are two branches (exocentric and egocentric) that need to be balanced and optimized during training. Moreover, the generalization performance of these models is suboptimal, as their supported classes are fixed (e.g, LOCATE (Li et al. 2023a)) or very few parameters are optimized on the affordance task (e.g., PLSP (Xu and Mu 2025)) . To tackle this problem, the second type of methods uses affordance maps to directly super-  \n∗Internship at SGIT AI Lab, State Grid Corporation of China.✉Corresponding author.†Equal contribution.  \nFigure 1: Performance Comparison: The circle area indicates the number of training data, with better-performing models positioned toward the upper right. Our AffordanceSAM and C2F-Aff data can respectively serve as an excellent base model and training data. Integrating the two achieves a performance far ahead of other candidates.  \nvise the models (Qian et al. 2024; Li et al. 2024) . However, some important affordance components trained from scratch (e.g., decoder) on only hundreds of pieces of manually labeled training data make it insufficient to obtain a highly generalized model. Thus, it is important for the fully supervised affordance grounding family to find a strong foundation model that is naturally suitable for the affordance grounding task and scale up the supervised data to obtain the ideal generalized model.  \nIn this work, we try to decompose and address this problem in a step-by-step manner:  \n(i) Incorporating a suitable and generalized vision foundation model. With the recent advancements in large-scale pre-trained foundation models, many works have attempted to leverage the prior knowledge for downstream task transfer (Wang et al. 2023a; Han and Lim 2024) . We believe this paradigm is also effective for the affordance grounding task. Given that affordance gro","cbCaibNdJga7rKXv","https://ap.wps.com/l/cbCaibNdJga7rKXv","pdf",3580637,1,13,"English","en",105,"# Abstract\n# Introduction\n## Affordance grounding and existing supervision paradigms\n## Using SAM as a vision foundation model\n## Transferring SAM capabilities via modules and data","[{\"question\":\"What problem does AffordanceSAM address in affordance grounding?\",\"answer\":\"It targets the difficulty of building a generalized affordance grounding model that can reliably identify actionable regions on objects for real-world use.\"},{\"question\":\"How does the method relate segmentation (SAM) to affordance grounding?\",\"answer\":\"It leverages SAM’s generalization and pixel-level localization strengths, noting structural similarity and guiding the decoder to refine masks toward affordance heatmaps based on verb queries.\"},{\"question\":\"What are the key components proposed in the paper?\",\"answer\":\"An affordance-adaption module with learnable queries and a coarse-to-fine annotated dataset called C2F-Aff, trained through a three-stage process to transfer SAM performance.\"}]",1784173997,33,{"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":85,"head_meta":87,"extra_data":89,"updated_unix":27},"affordancesam-segment-anything-once-more-in-affordance-grounding","",{"@graph":35,"@context":84},[36,53,67],{"@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/affordancesam-segment-anything-once-more-in-affordance-grounding/81524/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does AffordanceSAM address in affordance grounding?","Question",{"text":74,"@type":75},"It targets the difficulty of building a generalized affordance grounding model that can reliably identify actionable regions on objects for real-world use.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does the method relate segmentation (SAM) to affordance grounding?",{"text":79,"@type":75},"It leverages SAM’s generalization and pixel-level localization strengths, noting structural similarity and guiding the decoder to refine masks toward affordance heatmaps based on verb queries.",{"name":81,"@type":72,"acceptedAnswer":82},"What are the key components proposed in the paper?",{"text":83,"@type":75},"An affordance-adaption module with learnable queries and a coarse-to-fine annotated dataset called C2F-Aff, trained through a three-stage process to transfer SAM performance.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"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":105,"slug":137},19,"General","general"]