We propose RaysUp, an ultra-lightweight, task-agnostic, and VFM-agnostic upsampling framework, capable of upsampling backbone features to arbitrary resolutions while preserving high semantic fidelity and geometric consistency.</p>\n","updatedAt":"2026-06-30T12:03:23.145Z","author":{"_id":"69646cea2d94e9a07cdc90a9","avatarUrl":"/avatars/60cc696df7e505e26e141702d7b96096.svg","fullname":"kangkang","name":"DYC13145","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8567230105400085},"editors":["DYC13145"],"editorAvatarUrls":["/avatars/60cc696df7e505e26e141702d7b96096.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.22749","authors":[{"_id":"6a3bbb455ac9fb0744984967","user":{"_id":"69646cea2d94e9a07cdc90a9","avatarUrl":"/avatars/60cc696df7e505e26e141702d7b96096.svg","isPro":false,"fullname":"kangkang","user":"DYC13145","type":"user","name":"DYC13145"},"name":"Yuchuan Ding","status":"claimed_verified","statusLastChangedAt":"2026-06-25T09:28:54.914Z","hidden":false},{"_id":"6a3bbb455ac9fb0744984968","name":"Linfei Li","hidden":false},{"_id":"6a3bbb455ac9fb0744984969","name":"Lin Zhang","hidden":false},{"_id":"6a3bbb455ac9fb074498496a","name":"Ying Shen","hidden":false}],"publishedAt":"2026-06-22T00:00:00.000Z","submittedOnDailyAt":"2026-06-30T00:00:00.000Z","title":"RaysUp: Ultra-light Universal Feature Upsampling via Geometry-Aware Ray Representation","submittedOnDailyBy":{"_id":"69646cea2d94e9a07cdc90a9","avatarUrl":"/avatars/60cc696df7e505e26e141702d7b96096.svg","isPro":false,"fullname":"kangkang","user":"DYC13145","type":"user","name":"DYC13145"},"summary":"Pre-trained Vision Foundation Models (VFMs) have become central to modern computer vision due to their powerful semantic representations and strong generalization ability. However, their patchified or pooled outputs are inherently low-resolution, limiting their effectiveness in tasks requiring fine-grained, pixel-level reasoning. Existing feature upsampling approaches either degrade semantic fidelity or rely on VFM-specific retraining and heavy architectures, hindering efficiency and scalability. To address these challenges, we propose RaysUp, an ultra-lightweight, task-agnostic, and VFM-agnostic feature upsampling framework that reconstructs high-resolution feature maps at arbitrary resolutions. Unlike conventional 2D interpolation or attention-based schemes, RaysUp lifts feature reconstruction into a geometry-aware ray domain. Specifically, we introduce a Spatially Decoupled Guidance Encoder for direction-aware guidance encoding, an Any-Resolution Cross-Attention mechanism for resolution-flexible reconstruction, and a novel Ray Positional Encoding (RayPE) that injects implicit 3D geometric priors via 6D Plucker ray coordinates. Finally, a Geometry-Aware Neighborhood Attention module further ensures content-adaptive bilateral aggregation while preserving geometric consistency. Extensive experiments across diverse dense prediction tasks demonstrate that RaysUp achieves state-of-the-art performance while using only 16% of the parameters of AnyUp and delivering approximately 7x faster inference. These results highlight a substantially improved accuracy-efficiency trade-off and establish RaysUp as a practical and scalable solution for universal feature upsampling. Code is available at https://github.com/MAP-RaysUp/RaysUp.","upvotes":1,"discussionId":"6a3bbb455ac9fb074498496b","projectPage":"https://lif314.github.io/projects/raysup/","githubRepo":"https://github.com/MAP-RaysUp/RaysUp","githubRepoAddedBy":"user","ai_summary":"RaysUp is a lightweight, task-agnostic feature upsampling framework that reconstructs high-resolution features using geometry-aware ray domain techniques with improved efficiency and accuracy.","ai_keywords":["Vision Foundation Models","feature upsampling","spatially decoupled guidance encoder","any-resolution cross-attention","ray positional encoding","6D Plucker ray coordinates","geometry-aware neighborhood attention","dense prediction tasks","parameter efficiency","inference speed"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":10},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"69646cea2d94e9a07cdc90a9","avatarUrl":"/avatars/60cc696df7e505e26e141702d7b96096.svg","isPro":false,"fullname":"kangkang","user":"DYC13145","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.22749.md","query":{}}">
RaysUp: Ultra-light Universal Feature Upsampling via Geometry-Aware Ray Representation
Abstract
RaysUp is a lightweight, task-agnostic feature upsampling framework that reconstructs high-resolution features using geometry-aware ray domain techniques with improved efficiency and accuracy.
Pre-trained Vision Foundation Models (VFMs) have become central to modern computer vision due to their powerful semantic representations and strong generalization ability. However, their patchified or pooled outputs are inherently low-resolution, limiting their effectiveness in tasks requiring fine-grained, pixel-level reasoning. Existing feature upsampling approaches either degrade semantic fidelity or rely on VFM-specific retraining and heavy architectures, hindering efficiency and scalability. To address these challenges, we propose RaysUp, an ultra-lightweight, task-agnostic, and VFM-agnostic feature upsampling framework that reconstructs high-resolution feature maps at arbitrary resolutions. Unlike conventional 2D interpolation or attention-based schemes, RaysUp lifts feature reconstruction into a geometry-aware ray domain. Specifically, we introduce a Spatially Decoupled Guidance Encoder for direction-aware guidance encoding, an Any-Resolution Cross-Attention mechanism for resolution-flexible reconstruction, and a novel Ray Positional Encoding (RayPE) that injects implicit 3D geometric priors via 6D Plucker ray coordinates. Finally, a Geometry-Aware Neighborhood Attention module further ensures content-adaptive bilateral aggregation while preserving geometric consistency. Extensive experiments across diverse dense prediction tasks demonstrate that RaysUp achieves state-of-the-art performance while using only 16% of the parameters of AnyUp and delivering approximately 7x faster inference. These results highlight a substantially improved accuracy-efficiency trade-off and establish RaysUp as a practical and scalable solution for universal feature upsampling. Code is available at https://github.com/MAP-RaysUp/RaysUp.
Community
We propose RaysUp, an ultra-lightweight, task-agnostic, and VFM-agnostic upsampling framework, capable of upsampling backbone features to arbitrary resolutions while preserving high semantic fidelity and geometric consistency.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2606.22749 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.22749 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2606.22749 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.