Recent work on image content manipulation based on vision-language pre-training models has been effectively extended to text-driven 3D scene editing. However, existing schemes for 3D scene editing still have certain shortcomings, hindering their further development as interactive design tools. Such schemes typically adhere to fixed input patterns, limiting flexibility in text input. Furthermore, their editing capabilities are constrained by a single or a few 2D visual models and require intricate pipeline design to integrate these models into 3D reconstruction processes. To address the aforementioned issues, we propose the Hash-Atlas network, which reformulates 3D scene editing as operations on 2D atlas images, thereby achieving a workflow decoupling of the 2D editing and 3D reconstruction processes. Building on this foundation, we introduce a dialogue-based 3D scene editing approach, termed CE3D++, which is centered on a large language model (LLM) that allows arbitrary textual input from users and interprets their intentions, subsequently facilitating the autonomous invocation of the corresponding visual models. Additionally, we extend CE3D++ to monocular 4D scenes by imposing motion constraints on moving objects and further fine-tuning the LLM by creating a trajectory dataset related to editing tasks, which enables the smaller LLM to schedule up to 30 different visual tools accurately. Experimental results demonstrate that CE3D++ effectively integrates multiple visual models to achieve diverse visual editing effects, possessing strong scene comprehension and multi-round dialog capabilities. Code: <a href=\"https://sk-fun.fun/CE3D\" rel=\"nofollow\">https://sk-fun.fun/CE3D</a></p>\n","updatedAt":"2026-09-01T08:52:37.619Z","author":{"_id":"66ee3663518ccb4e62be37c4","avatarUrl":"/avatars/8d423c9bf63af13969300711a9efa870.svg","fullname":"Shuangkang Fang","name":"fsk515","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9017530679702759},"editors":["fsk515"],"editorAvatarUrls":["/avatars/8d423c9bf63af13969300711a9efa870.svg"],"reactions":[{"reaction":"🚀","users":["fsk515"],"count":1}],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.29137","authors":[{"_id":"6a969108cd6ebc484732efac","name":"Shuangkang Fang","hidden":false},{"_id":"6a969108cd6ebc484732efad","name":"Yufeng Wang","hidden":false},{"_id":"6a969108cd6ebc484732efae","name":"Yi-Hsuan Tsai","hidden":false},{"_id":"6a969108cd6ebc484732efaf","name":"Wenrui Ding","hidden":false},{"_id":"6a969108cd6ebc484732efb0","name":"Yi Yang","hidden":false},{"_id":"6a969108cd6ebc484732efb1","name":"Shuchang Zhou","hidden":false},{"_id":"6a969108cd6ebc484732efb2","name":"Ming-Hsuan Yang","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/66ee3663518ccb4e62be37c4/Hl7oBY8W-VunTOinYmTkp.mp4"],"publishedAt":"2026-08-29T00:00:00.000Z","submittedOnDailyAt":"2026-09-01T00:00:00.000Z","title":"Chat-Edit-3D++: Interactive 3D and 4D Scene Editing via Large Language Models","submittedOnDailyBy":{"_id":"66ee3663518ccb4e62be37c4","avatarUrl":"/avatars/8d423c9bf63af13969300711a9efa870.svg","isPro":false,"fullname":"Shuangkang Fang","user":"fsk515","type":"user","name":"fsk515"},"summary":"Recent work on image content manipulation based on vision-language pre-training models has been effectively extended to text-driven 3D scene editing. However, existing schemes for 3D scene editing still have certain shortcomings, hindering their further development as interactive design tools. Such schemes typically adhere to fixed input patterns, limiting flexibility in text input. Furthermore, their editing capabilities are constrained by a single or a few 2D visual models and require intricate pipeline design to integrate these models into 3D reconstruction processes. To address the aforementioned issues, we propose the Hash-Atlas network, which reformulates 3D scene editing as operations on 2D atlas images, thereby achieving a workflow decoupling of the 2D editing and 3D reconstruction processes. Building on this foundation, we introduce a dialogue-based 3D scene editing approach, termed CE3D++, which is centered on a large language model (LLM) that allows arbitrary textual input from users and interprets their intentions, subsequently facilitating the autonomous invocation of the corresponding visual models. Additionally, we extend CE3D++ to monocular 4D scenes by imposing motion constraints on moving objects and further fine-tuning the LLM by creating a trajectory dataset related to editing tasks, which enables the smaller LLM to schedule up to 30 different visual tools accurately. Experimental results demonstrate that CE3D++ effectively integrates multiple visual models to achieve diverse visual editing effects, possessing strong scene comprehension and multi-round dialog capabilities. The source codes and trained models are available at https://github.com/Fangkang515/CE3D.","upvotes":1,"discussionId":"6a969108cd6ebc484732efb3","projectPage":"https://sk-fun.fun/CE3D/","ai_summary":"CE3D++ enables flexible dialogue-driven 3D and 4D scene editing by decoupling 2D editing from 3D reconstruction via a Hash-Atlas network and using a large language model to autonomously invoke visual tools.","ai_keywords":["vision-language pre-training","Hash-Atlas network","3D scene editing","dialogue-based editing","large language model","monocular 4D scenes","motion constraints","trajectory dataset","visual models"],"ai_summary_model":"thinkingmachines/Inkling-Small","organization":{"_id":"66e43eae9d477f566f937935","name":"stepfun-ai","fullname":"StepFun","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/644f7e6233ac8f46fa0b9e26/CmF2ocXhkr2UtHXgmwq7-.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"66ee3663518ccb4e62be37c4","avatarUrl":"/avatars/8d423c9bf63af13969300711a9efa870.svg","isPro":false,"fullname":"Shuangkang Fang","user":"fsk515","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"66e43eae9d477f566f937935","name":"stepfun-ai","fullname":"StepFun","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/644f7e6233ac8f46fa0b9e26/CmF2ocXhkr2UtHXgmwq7-.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.29137.md","query":{}}">
Chat-Edit-3D++: Interactive 3D and 4D Scene Editing via Large Language Models
Abstract
CE3D++ enables flexible dialogue-driven 3D and 4D scene editing by decoupling 2D editing from 3D reconstruction via a Hash-Atlas network and using a large language model to autonomously invoke visual tools.
Recent work on image content manipulation based on vision-language pre-training models has been effectively extended to text-driven 3D scene editing. However, existing schemes for 3D scene editing still have certain shortcomings, hindering their further development as interactive design tools. Such schemes typically adhere to fixed input patterns, limiting flexibility in text input. Furthermore, their editing capabilities are constrained by a single or a few 2D visual models and require intricate pipeline design to integrate these models into 3D reconstruction processes. To address the aforementioned issues, we propose the Hash-Atlas network, which reformulates 3D scene editing as operations on 2D atlas images, thereby achieving a workflow decoupling of the 2D editing and 3D reconstruction processes. Building on this foundation, we introduce a dialogue-based 3D scene editing approach, termed CE3D++, which is centered on a large language model (LLM) that allows arbitrary textual input from users and interprets their intentions, subsequently facilitating the autonomous invocation of the corresponding visual models. Additionally, we extend CE3D++ to monocular 4D scenes by imposing motion constraints on moving objects and further fine-tuning the LLM by creating a trajectory dataset related to editing tasks, which enables the smaller LLM to schedule up to 30 different visual tools accurately. Experimental results demonstrate that CE3D++ effectively integrates multiple visual models to achieve diverse visual editing effects, possessing strong scene comprehension and multi-round dialog capabilities. The source codes and trained models are available at https://github.com/Fangkang515/CE3D.
Community
Recent work on image content manipulation based on vision-language pre-training models has been effectively extended to text-driven 3D scene editing. However, existing schemes for 3D scene editing still have certain shortcomings, hindering their further development as interactive design tools. Such schemes typically adhere to fixed input patterns, limiting flexibility in text input. Furthermore, their editing capabilities are constrained by a single or a few 2D visual models and require intricate pipeline design to integrate these models into 3D reconstruction processes. To address the aforementioned issues, we propose the Hash-Atlas network, which reformulates 3D scene editing as operations on 2D atlas images, thereby achieving a workflow decoupling of the 2D editing and 3D reconstruction processes. Building on this foundation, we introduce a dialogue-based 3D scene editing approach, termed CE3D++, which is centered on a large language model (LLM) that allows arbitrary textual input from users and interprets their intentions, subsequently facilitating the autonomous invocation of the corresponding visual models. Additionally, we extend CE3D++ to monocular 4D scenes by imposing motion constraints on moving objects and further fine-tuning the LLM by creating a trajectory dataset related to editing tasks, which enables the smaller LLM to schedule up to 30 different visual tools accurately. Experimental results demonstrate that CE3D++ effectively integrates multiple visual models to achieve diverse visual editing effects, possessing strong scene comprehension and multi-round dialog capabilities. Code: https://sk-fun.fun/CE3D
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/2608.29137 in a model README.md to link it from this page.
Cite arxiv.org/abs/2608.29137 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2608.29137 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.