[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86512-en":3,"doc-seo-86512-105":30,"detail-sidebar-cat-0-en-105":92},{"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},86512,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",8,"Research & Report","MAC-Splat Multi-Attribute Consistency for High-Fidelity Sparse-View Reconstruction","Reconstructing high-fidelity 3D scenes from sparse views remains a central problem in generalizable neural rendering. Existing generalizable 3D Gaussian Splatting methods often produce geometric artifacts under sparse-view settings because 2D photometric losses cannot disambiguate depth and correspondence. MAC-Splat introduces direct 3D consistency supervision: it uses MASt3R to derive semantically informed correspondences with a frozen DINOv3 encoder, then applies a Multi-Attribute Consistency loss to jointly regularize 3D position, shape, and appearance in a shared world frame. Experiments on ScanNet++ show strong gains, including over Splatt3R PSNR, lower LPIPS, and robustness as camera pose gaps increase.","arXiv :2607 . 10792v1 [ cs .CV] 12 Jul 2026  \nMAC-Splat: Multi-Attribute Consistency for High-Fidelity Sparse-View Reconstruction  \nJinqian Yang 1 , Yichen Wu2 * , Wanhua Li3 , Haokun Lin4 , Renzhen Wang5 , Xiangchu Feng 1 , and Xixi Jia 1 *  \n1 Xidian University, China  \n2 Harvard University, USA  \n3 Nanyang Technological University, Singapore  \n4 City University of Hong Kong, China  \n5 Xi’an Jiaotong University, China  \nAbstract. Reconstructing high-fidelity 3D scenes from sparse views remains a central problem in generalizable neural rendering. Existing generalizable 3D Gaussian Splatting (3DGS) methods often exhibit geometric artifacts in sparse-view settings, since supervision based solely on 2D photometric losses cannot resolve depth and correspondence ambiguities. To address this issue, we propose MAC-Splat, a training framework built around direct 3D consistency supervision. MAC-Splat builds on the MASt3R geometric backbone and a frozen DINOv3 encoder to obtain semantically informed 2D correspondences, which serve as geometric anchors for 3D supervision. Using these anchors, we define the MultiAttribute Consistency (MAC) loss. This objective jointly regularizes the 3D attributes of matched Gaussians, including their position, shape, and appearance, by enforcing agreement in a common world coordinate frame. The formulation is robust to outliers and respects the geometry of covariance matrices, which leads to stable training under sparse-view conditions. Experiments on ScanNet++ show that MAC-Splat outperforms strong baselines, with particularly large gains under different overlap regimes. In particular, it improves average PSNR over Splatt3R by more than 4.5 dB, reduces LPIPS, and maintains performance as the camera pose gap increases. These results indicate that a direct, multi-attribute 3D consistency objective, when combined with high-quality correspondences, is effective for addressing the ill-posed sparse-view reconstruction problem.  \nKeywords: 3D Gaussian Splatting · Sparse-View Reconstruction · Geometric Consistency · Semantic Guidance  \n1 Introduction  \nCapturing and rendering photorealistic 3D scenes is a central goal in computer vision [3, 37, 51, 57], with applications ranging from virtual reality [43] to  \n* Corresponding authors: Yichen Wu ([wuyichen.am97@gmail.com](wuyichen.am97@gmail.com)) and Xixi Jia ([xxjia@xidian.edu.cn](xxjia@xidian.edu.cn)).  \n2 J. Yang et al.  \nFig. 1: Matched-pixel constraints for 3D Gaussians. Pixel correspondences are used as anchors to enforce consistent Gaussian attributes, which helps mitigate common sparse-view artifacts compared to prior generalizable 3DGS methods.  \nrobotics [41] . Implicit neural representations such as NeRF [1,32] have demonstrated striking visual quality; however, their training and inference costs remain high, which limits real-time deployment and scalability in practical systems.  \nTo address this limitation, 3DGS and its variants [13, 20, 25, 36, 63] offer a powerful solution. They achieve real-time, high-quality rendering by using an explicit representation and an efficient GPU rasterization pipeline [49] . However, the strong performance of 3DGS usually relies on dense multi-view inputs with sufficient view overlap. In practice, when only a handful of images are available, reconstruction must proceed under sparse-view conditions, often with wide baselines and limited overlap [2,9,16,21,38,44,48,52,53,56,60,67,68] . This scarcity reduces geometric cues and thus impedes the formation of coherent 3D structures. Therefore, generalizing 3DGS to sparse-view regimes has become a key and rapidly evolving research focus [7,29] .  \nDespite the importance of this problem, existing mitigation strategies remain inadequate [7,61] . First, methods that inject geometric priors from depth estimators [14, 22, 30, 45] or leverage foundation models for indirect perceptual guidance [42,46] seldom provide pixel-level, cross-view anchors that directly tie corresponding ","cbCaio91eqGCST3y","https://ap.wps.com/l/cbCaio91eqGCST3y","pdf",2784323,6,1,19,"English","en",105,"# Introduction\n## Problem: sparse-view ambiguity and artifacts\n## Limitations of existing strategies\n## Proposed solution: MAC-Splat and MAC loss","[{\"question\":\"Why do existing generalizable 3D Gaussian Splatting methods struggle in sparse-view reconstruction?\",\"answer\":\"Because supervision based only on 2D photometric losses cannot resolve depth and correspondence ambiguities, leading to geometric artifacts when overlap is limited.\"},{\"question\":\"What is the core idea behind MAC-Splat?\",\"answer\":\"MAC-Splat complements 2D rendering losses with correspondence-grounded 3D consistency constraints, enforcing agreement across 3D position, shape, and appearance using semantically informed 2D correspondences as anchors.\"},{\"question\":\"How does MAC-Splat obtain the correspondences used for 3D supervision?\",\"answer\":\"It builds on the MASt3R geometric backbone and a frozen DINOv3 encoder to obtain semantically informed 2D correspondences, which serve as geometric anchors for the MAC loss.\"}]",1784212298,48,{"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":87,"head_meta":89,"extra_data":91,"updated_unix":28},"mac-splat-multi-attribute-consistency-for-high-fidelity-sparse-view-reconstruction","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"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":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/mac-splat-multi-attribute-consistency-for-high-fidelity-sparse-view-reconstruction/86512/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-07-24","2026-07-16",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why do existing generalizable 3D Gaussian Splatting methods struggle in sparse-view reconstruction?","Question",{"text":76,"@type":77},"Because supervision based only on 2D photometric losses cannot resolve depth and correspondence ambiguities, leading to geometric artifacts when overlap is limited.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What is the core idea behind MAC-Splat?",{"text":81,"@type":77},"MAC-Splat complements 2D rendering losses with correspondence-grounded 3D consistency constraints, enforcing agreement across 3D position, shape, and appearance using semantically informed 2D correspondences as anchors.",{"name":83,"@type":74,"acceptedAnswer":84},"How does MAC-Splat obtain the correspondences used for 3D supervision?",{"text":85,"@type":77},"It builds on the MASt3R geometric backbone and a frozen DINOv3 encoder to obtain semantically informed 2D correspondences, which serve as geometric anchors for the MAC loss.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,111,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},"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":22,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":107,"slug":137},"General","general"]