ShutterMuse is a unified multimodal large language model for capture-time photography guidance. It supports:</p>\n<ul>\n<li>Photographer-side guidance: keep, refine, or reject the current framing, with a composition box when refinement is needed.</li>\n<li>Subject-side guidance: recommend scene-conditioned portrait poses with COCO-17 keypoints and visibility states.</li>\n</ul>\n","updatedAt":"2026-06-25T06:57:25.117Z","author":{"_id":"647469b9a51711a3b58bda2b","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/647469b9a51711a3b58bda2b/yeDf8Sa8IDEQyney1dGC9.jpeg","fullname":"Yixiao Fang","name":"fangyixiao","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9065771102905273},"editors":["fangyixiao"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/647469b9a51711a3b58bda2b/yeDf8Sa8IDEQyney1dGC9.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.25763","authors":[{"_id":"6a3cb91cf3facdb67e9ff240","name":"Jiayu Li","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff241","name":"Yixiao Fang","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff242","name":"Tianyu Hu","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff243","name":"Wei Cheng","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff244","name":"Ping Huang","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff245","name":"Zheheng Fan","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff246","name":"Gang Yu","hidden":false},{"_id":"6a3cb91cf3facdb67e9ff247","name":"Xingjun Ma","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/647469b9a51711a3b58bda2b/H971Df6p2IGq8gXpvjuJG.mp4"],"publishedAt":"2026-06-24T00:00:00.000Z","submittedOnDailyAt":"2026-06-25T00:00:00.000Z","title":"ShutterMuse: Capture-Time Photography Guidance with MLLMs","submittedOnDailyBy":{"_id":"647469b9a51711a3b58bda2b","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/647469b9a51711a3b58bda2b/yeDf8Sa8IDEQyney1dGC9.jpeg","isPro":false,"fullname":"Yixiao Fang","user":"fangyixiao","type":"user","name":"fangyixiao"},"summary":"Real-world photography requires capture-time guidance for both camera framing and subject pose. Yet existing aesthetic cropping benchmarks mainly evaluate post-hoc crop prediction and overlook subject-side recommendations, leaving the capture-time guidance capabilities of multimodal large language models (MLLMs) underexplored. To address this gap, we introduce CaptureGuide-Bench, a benchmark with two complementary tasks: photographer-side composition decision and refinement, and subject-side scene-conditioned pose recommendation. Our evaluation reveals limitations: general-purpose MLLMs can make composition decisions but lack precise refinement localization, while specialized aesthetic cropping models localize crops effectively but are limited to refinement; neither provides actionable pose guidance. To support model development, we further construct CaptureGuide-Dataset, comprising 130K samples with textual rationales and structured visual annotations, and develop ShutterMuse, a unified MLLM trained with supervised and reinforcement fine-tuning. Experiments on CaptureGuide-Bench show that ShutterMuse achieves the best overall photographer-side performance among evaluated baselines and competitive subject-side pose recommendation with substantially lower inference cost, demonstrating the potential of MLLMs as interactive assistants for photography during image capture.","upvotes":33,"discussionId":"6a3cb91cf3facdb67e9ff248","projectPage":"https://lijayutnt.github.io/ShutterMuse/","githubRepo":"https://github.com/lijayuTnT/ShutterMuse","githubRepoAddedBy":"user","ai_summary":"Researchers developed a new benchmark and dataset for photography assistance, along with a unified multimodal model that provides both composition guidance and pose recommendations during image capture.","ai_keywords":["multimodal large language models","aesthetic cropping","visual annotations","supervised fine-tuning","reinforcement fine-tuning","photographer-side composition","subject-side pose recommendation"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":11},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6a26a58972e3d7254eb93bf8","avatarUrl":"/avatars/f16576a32a62b4c66ff1741411e2174c.svg","isPro":false,"fullname":"Lee","user":"ShutterMuse","type":"user"},{"_id":"647469b9a51711a3b58bda2b","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/647469b9a51711a3b58bda2b/yeDf8Sa8IDEQyney1dGC9.jpeg","isPro":false,"fullname":"Yixiao Fang","user":"fangyixiao","type":"user"},{"_id":"64b914c8ace99c0723ad83a9","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64b914c8ace99c0723ad83a9/B4gxNByeVY_xaOcjwiN1j.jpeg","isPro":false,"fullname":"Wei Cheng","user":"wchengad","type":"user"},{"_id":"669dcdb39a4bf63e08f70e9f","avatarUrl":"/avatars/1ecaf09e4c4f5355ea25a0606ff8f7e2.svg","isPro":false,"fullname":"yulatu","user":"yulatu","type":"user"},{"_id":"6343de25e01a38440ef02d5e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6343de25e01a38440ef02d5e/eumgZKT6vfTzINC6cYUrL.jpeg","isPro":false,"fullname":"xz","user":"frankzeng","type":"user"},{"_id":"6675854966c4fa6d0cee4d50","avatarUrl":"/avatars/aa6041a97985078e82cc89bfbade9828.svg","isPro":false,"fullname":"xuanyang zhang","user":"xuanyangz","type":"user"},{"_id":"64ca05b4f7f4ccb5ea6e43aa","avatarUrl":"/avatars/c909613715eaf5fd43ae6cd95ae2b9a4.svg","isPro":false,"fullname":"Charles","user":"SCFW","type":"user"},{"_id":"66cf5df9a181ad5423f6a3fb","avatarUrl":"/avatars/72b85e87e55865ba389b5015d2e20ea7.svg","isPro":false,"fullname":"Shuhan Wu","user":"LizzyWu","type":"user"},{"_id":"68fa28bd3ccb95f78051210e","avatarUrl":"/avatars/bf3e04e78cd48c10f24341cfc96126c7.svg","isPro":false,"fullname":"linghuyuhangyuan","user":"linghuyuhangyuan123","type":"user"},{"_id":"6555c84691e52f423173def5","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6555c84691e52f423173def5/fzD9xvDWy2MYVGAqmTC2I.png","isPro":false,"fullname":"leixinlin","user":"leixinlin","type":"user"},{"_id":"655101623fe6c0b1f8b58987","avatarUrl":"/avatars/4d36a4988e6011fec3ceac2b59938c3a.svg","isPro":false,"fullname":"Jiabin Hua","user":"Ammmob","type":"user"},{"_id":"62f361e6231737ed2d741740","avatarUrl":"/avatars/f4a1053f9d9b3e703d138bc9753742c1.svg","isPro":false,"fullname":"huyaoqi","user":"yaoqi","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":3,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.25763.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

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