[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81611-en":3,"doc-seo-81611-105":30,"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":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},81611,34359740700684,"Finn","https://ap-avatar.wpscdn.com/avatar/1f400023980c374ae676?_k=1777273430885731487",8,"Research & Report","Any to Full: Prompting Depth Anything for Depth Completion in One Stage","Accurate dense depth estimation is critical for robotic perception, yet commodity sensors often provide sparse or incomplete depth due to hardware limits. Existing RGBD-fused completion methods learn priors tied to training RGB distributions and specific depth patterns, hurting domain generalization and robustness. Monocular depth estimation can help, but current two-stage integrations require explicit relative-to-metric alignment, increasing compute and causing structured distortions. Any2Full proposes a one-stage, domain-general, pattern-agnostic approach using scale-prompting adaptation with a Scale-Aware Prompt Encoder.","arXiv :2603 .05711v2 [ cs .CV] 10 Jul 2026  \nAny to Full: Prompting Depth Anything for Depth Completion in One Stage  \nZhiyuan Zhou 1 , Ruofeng Liu2 B , Taichi Liu 1 , Weijian Zuo3 , Shanshan Wang 1 , Zhiqing Hong4 , and Desheng Zhang 1  \n1 Rutgers University, USA {zhiyuan.z,taichi.liu,[shanshan.wang}@rutgers.edu](shanshan.wang}@rutgers.edu) , [desheng@cs.rutgers.edu](desheng@cs.rutgers.edu)  \n2 Michigan State University, USA [liuruofe@msu.edu](liuruofe@msu.edu)  \n3 JD Logistics, China [zuoweijian1@jd.com](zuoweijian1@jd.com)  \n4 HKUST (Guangzhou), China [zhiqinghong@hkust-gz.edu.cn](zhiqinghong@hkust-gz.edu.cn)  \nAbstract. Accurate, dense depth estimation is crucial for robotic perception, but commodity sensors often yield sparse or incomplete measurements due to hardware limitations. Existing RGBD-fused depth completion methods learn priors jointly conditioned on training RGB distribution and specific depth patterns, limiting domain generalization and robustness to various depth patterns. Recent efforts leverage monocular depth estimation (MDE) models to introduce domain-general geometric priors, but current two-stage integration strategies relying on explicit relative-to-metric alignment incur additional computation and introduce structured distortions. To this end, we present Any2Full, a one-stage, domain-general, and pattern-agnostic framework that reformulates completion as a scale-prompting adaptation of a pretrained MDE model.  \nTo address varying depth sparsity levels and irregular spatial distributions, we design a Scale-Aware Prompt Encoder. It distills scale cues from sparse inputs into unified scale prompts, guiding the MDE model toward globally scale-consistent predictions while preserving its geometric priors. Extensive experiments demonstrate that Any2Full achieves superior robustness and efficiency. It outperforms OMNI-DC by 32 .2% in average AbsREL and delivers a 1.4× speedup over PriorDA with the same MDE backbone, establishing a new paradigm for universal depth completion. Codes and checkpoints are available at [https://github](https://github) .  \ncom/zhiyuandaily/Any2Full.  \n1 Introduction  \nAccurate and fine-grained depth information is essential for robotic perception, enabling navigation [38, 55], manipulation [14, 37, 52], and scene understanding [2, 7, 39] . However, commodity depth sensors (e.g., LiDAR [19], ToF [18], or structured light cameras [49]) often yield sparse or incomplete depth maps due  \nto limitations in resolution, range, and light interaction, such as reflection or absorption [6 1] . Consequently, depth completion has emerged as a fundamental B Corresponding author.  \n2 Z. Zhou et al.  \nFig. 1: Depth completion recovers dense depth maps from raw measurements and RGB guidance. (a) Traditional two-stage methods predict coarse depth to bridge the gap between sparse input and dense output. (b) Recent approaches integrate monocular depth estimation (MDE) to generate relative depth and explicitly align it with sparse depth, disrupting MDE’s geometric priors. (c) Our framework employs a lightweight prompting mechanism that tightly integrates MDE’s geometric priors, achieving domaingeneral and pattern-agnostic depth completion in one stage.  \ntask that aims to recover dense metric depth maps from raw depth measurements and their corresponding RGB images.  \nMost existing depth completion methods, as illustrated in Fig. 1(a), learn geometric priors from RGB images to guide dense depth prediction from sparse inputs [10, 25, 42, 47, 53, 54, 75] . However, as shown in Fig. 1(left), these RGBD fused frameworks, such as CompFormer [75], learn priors jointly conditioned on RGB distribution and specific depth patterns observed during training, leading to two commonly observed limitations: (1) Domain Specificity, where performance degrades under visual domain shifts such as lighting, texture, or scene variations. (2) Depth Pattern Sensitivity, where performance degrades with changing raw depth patterns","cbCaikr38kdh6DDp","https://ap.wps.com/l/cbCaikr38kdh6DDp","pdf",23417900,3,1,32,"English","en",105,"# Abstract\n# Introduction\n## Problem: sparse depth from commodity sensors\n## Limitations of RGBD fused and two-stage MDE integration\n## Proposed solution: Any2Full one-stage scale-prompting framework","[{\"question\":\"What problem does Any2Full target in depth completion?\",\"answer\":\"It addresses inaccurate dense depth estimation when depth sensors produce sparse or incomplete measurements, aiming to recover dense metric depth using RGB guidance more robustly across domains and depth patterns.\"},{\"question\":\"Why do existing RGBD-fused and MDE-based methods struggle with robustness?\",\"answer\":\"RGBD-fused approaches rely on priors conditioned on training RGB distributions and particular depth patterns, while current MDE integrations often use two-stage explicit alignment that introduces additional computation and structured scale distortions.\"},{\"question\":\"What is the key idea behind Any2Full?\",\"answer\":\"Any2Full reformulates depth completion as a one-stage scale-prompting adaptation of a pretrained monocular depth estimation model, using a Scale-Aware Prompt Encoder to distill unified scale cues from sparse inputs.\"}]",1784174766,81,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"any-to-full-prompting-depth-anything-for-depth-completion-in-one-stage","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":20},"https://docshare.wps.com/document/research-report/",{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/any-to-full-prompting-depth-anything-for-depth-completion-in-one-stage/81611/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-25","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 Any2Full target in depth completion?","Question",{"text":75,"@type":76},"It addresses inaccurate dense depth estimation when depth sensors produce sparse or incomplete measurements, aiming to recover dense metric depth using RGB guidance more robustly across domains and depth patterns.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why do existing RGBD-fused and MDE-based methods struggle with robustness?",{"text":80,"@type":76},"RGBD-fused approaches rely on priors conditioned on training RGB distributions and particular depth patterns, while current MDE integrations often use two-stage explicit alignment that introduces additional computation and structured scale distortions.",{"name":82,"@type":73,"acceptedAnswer":83},"What is the key idea behind Any2Full?",{"text":84,"@type":76},"Any2Full reformulates depth completion as a one-stage scale-prompting adaptation of a pretrained monocular depth estimation model, using a Scale-Aware Prompt Encoder to distill unified scale cues from sparse inputs.","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":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]