[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84819-en":3,"doc-seo-84819-105":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},84819,2336464648322,"Aria","https://ap-avatar.wpscdn.com/avatar/2200025388227c56fec?_k=1778556882303663488",8,"Research & Report","PixWorld Unifying 3D Scene Generation and Reconstruction in Pixel Space","3D reconstruction and generation are often treated separately, with pixel-based regression handling reconstruction and latent diffusion focusing on generation. Prior unified approaches define diffusion on latent features, which cannot directly optimize the underlying 3D representation and introduces information loss from pretrained VAE/RAE encoders. PixWorld reformulates both tasks under a unified pixel-space diffusion framework, supervising diffusion directly on rendered images. It further adds a geometry perception loss using geometry-aware features from a pretrained 3D foundation model to provide 3D structural supervision.","arXiv :2607 .05373v 1 [ cs .CV] 6 Jul 2026  \nPIXWORLD: UNIFYING 3D SCENE GENERATION AND RECONSTRUCTION IN PIXEL SPACE  \nSensen Gao 1∗ , Zhaoqing Wang2∗, Qihang Cao 1 , Dongdong Yu2 , Changhu Wang2 , Jia-Wang Bian2†  \n1 Nanyang Technological University 2 AISphere  \n* Co-first authors. †Corresponding authors.  \nInput Images Training Objective (latent space) 3DGS  \n| \u003Cbr> | \u003Cbr>\u003Cbr>3DGS Rendered Images | Training Objective (3D Representation) |\n| --- | --- | --- |\n\nFigure 1: PixWorld unifies 3D scene reconstruction and generation within a single model. Unlike prior approaches that compute losses in the latent space of a VAE (Kingma & Welling, 2013) or RAE (Zheng et al., 2025), PixWorld applies a flow matching objective directly in pixel space over multi-view renderings, enabling end-to-end optimization of the underlying 3D representation. This design avoids the information loss inherent to latent representations and eliminates the cost of pretraining a VAE or RAE.  \nABSTRACT  \n3D reconstruction and generation are commonly tackled by separate paradigms: pixel-based regression for reconstruction, and latent diffusion for generation. Recent works attempt to unify them in latent space, but with notable drawbacks:  \nthe diffusion objective is defined on latent features rather than the underlying 3D representation, and both branches suffer from information loss introduced by latent encoding, while requiring a pretrained Variational Autoencoder (VAE) or Representation Autoencoder (RAE) . In this paper, we reformulate these two tasks under a unified pixel-space diffusion paradigm and introduce PixWorld, a single model that jointly addresses 3D reconstruction and generation. By supervising diffusion directly on rendered images, PixWorld removes the above limitations and aligns optimization with 3D scene fidelity. Beyond photometric and perceptual supervision that operates at the 2D image level and lacks 3D geometric awareness, we further introduce a geometry perception loss that aligns rendered views with their ground truth in the geometry-aware feature space of a pretrained 3D foundation model, providing 3D structural supervision. PixWorld consistently outperforms  \nprior latent-space generation methods and matches state-of-the-art reconstruction methods, demonstrating the superiority of a unified pixel-space approach.  \n1 INTRODUCTION  \nBuilding 3D scenes from visual observations is a long-standing goal in computer vision, with broad impact across gaming, robotics, embodied AI, and VR/AR (Ding et al., 2025; Kong et al., 2025; Ye et al., 2026b; Zhang et al., 2025) . Two complementary directions have driven progress in this space: reconstruction recovers 3D scenes from real-world captures, while generation synthesizes plausible scenes from limited or even imagined conditions (Zhang et al., 2025; Wang et al., 2026; Liet al., 2026) . Together, they form the technological foundation for populating the digital worlds of tomorrow.  \n3D scene generation and reconstruction have long developed as two separate lines of research. Reconstruction is dominated by feed-forward methods that regress 3D representations directly from multi-view images (Hong et al., 2023; Charatan et al., 2024; Szymanowicz et al., 2024; Xu et al., 2025; Chen et al., 2024a) . Generation has evolved from per-scene optimization with 2D priors via score distillation (Lin et al., 2023; Poole et al., 2022; Tang et al., 2023; Wang et al., 2023) to latent-space diffusion as the current mainstream (Yang et al., 2025; Go et al., 2025b; Huang et al., 2026; Li et al., 2025b), with recent extensions diffusing in the feature space of pretrained 3D foundation models or representation autoencoders (Gao et al., 2026; Sun et al., 2026; Jang et al., 2026) . A recent work, Gen3R (Huang et al., 2026), attempts to unify the two tasks by extending the latent-space generation pipeline to also handle reconstruction. However, this design introduces clear limitations: the diffusion objective is defin","cbCaiikDhSUFVIiN","https://ap.wps.com/l/cbCaiikDhSUFVIiN","pdf",21223986,1,22,"English","en",105,"# Introduction\n## Unifying reconstruction and generation\n## Limitations of latent-space unification\n## PixWorld pixel-space diffusion framework\n## Geometry perception loss","[{\"question\":\"What problem does PixWorld address in existing unified reconstruction and generation methods?\",\"answer\":\"Existing methods apply diffusion on latent features rather than directly optimizing the underlying 3D representation, and they suffer information loss from pretrained VAE or RAE encoders.\"},{\"question\":\"How does PixWorld unify 3D reconstruction and generation differently?\",\"answer\":\"PixWorld reformulates both tasks under a single pixel-space diffusion paradigm, supervising diffusion directly on rendered images to align optimization with 3D scene fidelity.\"},{\"question\":\"Why is a geometry perception loss introduced in PixWorld?\",\"answer\":\"Image-level photometric and perceptual supervision may not ensure geometrically faithful 3D structure, so the geometry perception loss aligns rendered views with ground truth in a geometry-aware feature space from a pretrained 3D foundation model.\"}]",1784198496,55,{"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},"pixworld-unifying-3d-scene-generation-and-reconstruction-in-pixel-space","",{"@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/pixworld-unifying-3d-scene-generation-and-reconstruction-in-pixel-space/84819/",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 PixWorld address in existing unified reconstruction and generation methods?","Question",{"text":75,"@type":76},"Existing methods apply diffusion on latent features rather than directly optimizing the underlying 3D representation, and they suffer information loss from pretrained VAE or RAE encoders.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does PixWorld unify 3D reconstruction and generation differently?",{"text":80,"@type":76},"PixWorld reformulates both tasks under a single pixel-space diffusion paradigm, supervising diffusion directly on rendered images to align optimization with 3D scene fidelity.",{"name":82,"@type":73,"acceptedAnswer":83},"Why is a geometry perception loss introduced in PixWorld?",{"text":84,"@type":76},"Image-level photometric and perceptual supervision may not ensure geometrically faithful 3D structure, so the geometry perception loss aligns rendered views with ground truth in a geometry-aware feature space from a pretrained 3D foundation model.","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},[93,97,101,105,110,115,120,123,128,131,135],{"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":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]