Scalable online 3D reconstruction on kilometer-scale sequences, with only ~1% extra parameters on a frozen backbone trained in 8 hours on a single GPU.</p>\n","updatedAt":"2026-09-04T08:15:36.668Z","author":{"_id":"666afb91e936f6cbcfc8b50c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/666afb91e936f6cbcfc8b50c/_lcbPagwDTn02TaOSDxUq.jpeg","fullname":"Chin-Yang Lin","name":"linjohnss","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false,"primaryOrg":{"avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65df9200dc3292a8983e5017/Vs5FPVCH-VZBipV3qKTuy.png","fullname":"NVIDIA","name":"nvidia","type":"org","isHf":false,"plan":"plus"}}},"numEdits":1,"identifiedLanguage":{"language":"en","probability":0.8701691627502441},"editors":["linjohnss"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/666afb91e936f6cbcfc8b50c/_lcbPagwDTn02TaOSDxUq.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2609.04201","authors":[{"_id":"6a9a7c688f7c3b75572396f2","name":"Chin-Yang Lin","hidden":false},{"_id":"6a9a7c688f7c3b75572396f3","name":"Yang-Che Sun","hidden":false},{"_id":"6a9a7c688f7c3b75572396f4","name":"Cheng Sun","hidden":false},{"_id":"6a9a7c688f7c3b75572396f5","name":"Fu-En Yang","hidden":false},{"_id":"6a9a7c688f7c3b75572396f6","name":"Min-Hung Chen","hidden":false},{"_id":"6a9a7c688f7c3b75572396f7","name":"Yen-Yu Lin","hidden":false},{"_id":"6a9a7c688f7c3b75572396f8","name":"Wei-Chen Chiu","hidden":false},{"_id":"6a9a7c688f7c3b75572396f9","name":"Yu-Lun Liu","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/666afb91e936f6cbcfc8b50c/Bm8udL0gcM3eoYFjD3ORl.mp4"],"publishedAt":"2026-09-03T00:00:00.000Z","submittedOnDailyAt":"2026-09-04T00:00:00.000Z","title":"Scal3R: Learning Efficient Multi-Relative Pose Query for Scalable Online 3D Reconstruction","submittedOnDailyBy":{"_id":"666afb91e936f6cbcfc8b50c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/666afb91e936f6cbcfc8b50c/_lcbPagwDTn02TaOSDxUq.jpeg","isPro":false,"fullname":"Chin-Yang Lin","user":"linjohnss","type":"user","name":"linjohnss"},"summary":"Online 3D reconstruction models perform poorly on long videos. This happens because regressing poses relative to a fixed first-frame anchor forces extrapolation far beyond the training distribution. Small drifts accumulate and amplify into significant geometric collapse. However, we observe that per-frame depth remains stable throughout this failure. The backbone's local geometry remains intact; only the global pose head breaks down. Motivated by this decoupling, we introduce Scal3R. This approach reformulates online reconstruction as multi-reference relative pose querying. We use lightweight learnable tokens, which make up about ~1% of the parameters, and inject them into a completely frozen backbone via asymmetric attention. This setup queries poses relative to multiple past keyframes. An online pose-graph optimization system with loop closure suppresses long-range drift. Scal3R reaches convergence in 8 hours on a single GPU. It reduces the average ATE by over 60% on KITTI compared to the online baseline. It also achieves state-of-the-art performance across Virtual KITTI, Sintel, TUM-Dynamic, ScanNet, and 7-Scenes. Project page: https://linjohnss.github.io/scal3r/","upvotes":21,"discussionId":"6a9a7c698f7c3b75572396fa","projectPage":"https://linjohnss.github.io/scal3r/","ai_summary":"Scal3R improves long-video online 3D reconstruction by querying multi-reference relative poses with lightweight tokens and pose-graph optimization, reducing drift without retraining the backbone.","ai_keywords":["online 3D reconstruction","relative pose querying","learnable tokens","asymmetric attention","frozen backbone","pose-graph optimization","loop closure","ATE"],"ai_summary_model":"thinkingmachines/Inkling-Small","organization":{"_id":"63e39e6499a032b1c950403d","name":"NYCU","fullname":"National Yang Ming Chiao Tung University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/63e39df6c65f975b436bb6b8/WLWf1bSpvrXBYYKEdXbgU.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6459d5da3b6fafd9664807ab","avatarUrl":"/avatars/57430d1bbde3a2fe5586e5fbcafb0e74.svg","isPro":false,"fullname":"Yu-Lun Liu","user":"yulunliu","type":"user"},{"_id":"65d61572531de9717708be9f","avatarUrl":"/avatars/58c30060e847ddaefe3e8f0a88008725.svg","isPro":false,"fullname":"Chih-hai Su","user":"ohinic","type":"user"},{"_id":"6672ebc506b6d49dda7598c5","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6672ebc506b6d49dda7598c5/9yUeKzZZVtBoy2L-dNPMf.png","isPro":false,"fullname":"Sytwu","user":"Sytwu","type":"user"},{"_id":"666afb91e936f6cbcfc8b50c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/666afb91e936f6cbcfc8b50c/_lcbPagwDTn02TaOSDxUq.jpeg","isPro":false,"fullname":"Chin-Yang Lin","user":"linjohnss","type":"user"},{"_id":"6672fe26c33b5004b69a1d6a","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/Ff8cOS6Y0TPUSihx_hOMe.png","isPro":false,"fullname":"YouZhe","user":"YouZhe","type":"user"},{"_id":"68a41489d9b513a884bca475","avatarUrl":"/avatars/e4caeb16f3c4e7c36835cf26c8cb0d2c.svg","isPro":false,"fullname":"You-Zhe Xie","user":"YouZhe0305","type":"user"},{"_id":"68e0130b12b8c1ecde3383ae","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/gptMaTL5NfK5gLqVH_1zm.png","isPro":false,"fullname":"You-Zhe Xie","user":"Beck0305","type":"user"},{"_id":"64ea1e12925565abda02b17b","avatarUrl":"/avatars/b2bc33d95a147c6c8cf6b54672eb5a97.svg","isPro":false,"fullname":"Cheng-De Fan","user":"fansam39","type":"user"},{"_id":"66040e5acfb9b90f95a31d76","avatarUrl":"/avatars/bd5b8a4c38aac0531126fd036debb5ed.svg","isPro":false,"fullname":"Zheng-Hui Huang","user":"Brian9999","type":"user"},{"_id":"64cdecee2f1f9578a0e701c8","avatarUrl":"/avatars/95a51dd4e1b7b9366ebcbd6028ad148b.svg","isPro":false,"fullname":"Yi-Ruei Liu","user":"Shigon","type":"user"},{"_id":"67178582bc4492cad19a1f14","avatarUrl":"/avatars/f2481c0c70a857a862d887beb05c428e.svg","isPro":false,"fullname":"Yi-Chuan Huang","user":"YiChuanH","type":"user"},{"_id":"6312cab05beb528b5c1500e3","avatarUrl":"/avatars/fc27d245dbf06a41c38aa8da880a02cb.svg","isPro":false,"fullname":"Fu-En Yang","user":"FuEnYang","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"63e39e6499a032b1c950403d","name":"NYCU","fullname":"National Yang Ming Chiao Tung University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/63e39df6c65f975b436bb6b8/WLWf1bSpvrXBYYKEdXbgU.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2609/2609.04201.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2609.04201 in a model README.md to link it from this page.

Cite arxiv.org/abs/2609.04201 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2609.04201 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.