3DarkFusion fuses extremely noisy RGB with Near-Infrared (NIR) in 3D space via neural rendering -- no clean RGB supervision needed, robust even under severe noise interference.</p>\n","updatedAt":"2026-08-03T08:48:20.192Z","author":{"_id":"66597e53085eb651879aa4a1","avatarUrl":"/avatars/116aa003266bb19e8e03eb7ac5633b03.svg","fullname":"Muyao Niu","name":"MyNiuuu","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":7,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7678782939910889},"editors":["MyNiuuu"],"editorAvatarUrls":["/avatars/116aa003266bb19e8e03eb7ac5633b03.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2607.29684","authors":[{"_id":"6a7055ad02c90f968f48a199","name":"Muyao Niu","hidden":false},{"_id":"6a7055ad02c90f968f48a19a","name":"Mingze Ma","hidden":false},{"_id":"6a7055ad02c90f968f48a19b","name":"Yifan Zhan","hidden":false},{"_id":"6a7055ad02c90f968f48a19c","name":"Qingtian Zhu","hidden":false},{"_id":"6a7055ad02c90f968f48a19d","name":"Zhihang Zhong","hidden":false},{"_id":"6a7055ad02c90f968f48a19e","name":"Wei Guo","hidden":false},{"_id":"6a7055ad02c90f968f48a19f","name":"Chang Wen Chen","hidden":false},{"_id":"6a7055ad02c90f968f48a1a0","name":"Yinqiang Zheng","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/66597e53085eb651879aa4a1/xArYGP4Xg08OqLOb4fz7T.gif"],"publishedAt":"2026-07-31T00:00:00.000Z","submittedOnDailyAt":"2026-08-03T00:00:00.000Z","title":"Toward Robust and 3D-Aware RGB-NIR Imaging in the Dark","submittedOnDailyBy":{"_id":"66597e53085eb651879aa4a1","avatarUrl":"/avatars/116aa003266bb19e8e03eb7ac5633b03.svg","isPro":false,"fullname":"Muyao Niu","user":"MyNiuuu","type":"user","name":"MyNiuuu"},"summary":"Robust low-light imaging remains challenging for the community. Recent studies have explored fusing Near-Infrared (NIR) with noisy RGB to achieve improved enhancement, yet most methods depend on carefully curated training data pairs, with limited robustness under different scenarios. This paper offers a new perspective for RGB-NIR low-light imaging by incorporating 3D-aware neural modeling. Without using clean RGB supervision, a powerful model can be optimized to implicitly fuse extremely noisy RGB observations with NIR cues in 3D space, effectively recovering clean RGB images. The proposed model obviates the requirement for clean RGB data collection, generalizes across different noise levels. Extensive evaluations on synthetic and real data demonstrate its superiority. Codes available: https://github.com/MyNiuuu/3DarkFusion","upvotes":2,"discussionId":"6a7055ad02c90f968f48a1a1","githubRepo":"https://github.com/MyNiuuu/3DarkFusion","githubRepoAddedBy":"user","githubStars":1},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"66597e53085eb651879aa4a1","avatarUrl":"/avatars/116aa003266bb19e8e03eb7ac5633b03.svg","isPro":false,"fullname":"Muyao Niu","user":"MyNiuuu","type":"user"},{"_id":"65c4eb7cd1dcbd30d86febec","avatarUrl":"/avatars/001c8f02e8ce794b2c21883628b2da72.svg","isPro":false,"fullname":"free-bit","user":"free-bit","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2607/2607.29684.md","query":{}}">
Toward Robust and 3D-Aware RGB-NIR Imaging in the Dark
Abstract
Robust low-light imaging remains challenging for the community. Recent studies have explored fusing Near-Infrared (NIR) with noisy RGB to achieve improved enhancement, yet most methods depend on carefully curated training data pairs, with limited robustness under different scenarios. This paper offers a new perspective for RGB-NIR low-light imaging by incorporating 3D-aware neural modeling. Without using clean RGB supervision, a powerful model can be optimized to implicitly fuse extremely noisy RGB observations with NIR cues in 3D space, effectively recovering clean RGB images. The proposed model obviates the requirement for clean RGB data collection, generalizes across different noise levels. Extensive evaluations on synthetic and real data demonstrate its superiority. Codes available: https://github.com/MyNiuuu/3DarkFusion
Community
3DarkFusion fuses extremely noisy RGB with Near-Infrared (NIR) in 3D space via neural rendering -- no clean RGB supervision needed, robust even under severe noise interference.
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/2607.29684 in a model README.md to link it from this page.
Cite arxiv.org/abs/2607.29684 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2607.29684 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.