https://tailor3d-2024.github.io/\n","updatedAt":"2024-07-09T06:08:03.989Z","author":{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","fullname":"Jiaqi Wang","name":"myownskyW7","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":25,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.5444436073303223},"editors":["myownskyW7"],"editorAvatarUrls":["/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg"],"reactions":[{"reaction":"🔥","users":["AdinaY"],"count":1}],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2407.06191","authors":[{"_id":"668cb6cc61b6eff5a82aeba2","user":{"_id":"64c129c155686a357c99ea0b","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64c129c155686a357c99ea0b/1iy8_aLVU4FflO-d1fozO.jpeg","isPro":true,"fullname":"Qi Zhangyang (Alex Severns)","user":"alexzyqi","type":"user"},"name":"Zhangyang Qi","status":"admin_assigned","statusLastChangedAt":"2024-07-09T14:05:53.230Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba3","user":{"_id":"64b4eecf2fc8324fcb63b404","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64b4eecf2fc8324fcb63b404/zGYqYVB4-o-GBMybJ8CDA.png","isPro":false,"fullname":"Yunhan Yang","user":"yhyang-myron","type":"user"},"name":"Yunhan Yang","status":"admin_assigned","statusLastChangedAt":"2024-07-09T14:06:02.450Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba4","user":{"_id":"64de20c5808492ba6e65d124","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64de20c5808492ba6e65d124/58IX_TI5vJw73qS1knw56.jpeg","isPro":false,"fullname":"Zhang Mengchen","user":"Dubhe-zmc","type":"user"},"name":"Mengchen Zhang","status":"admin_assigned","statusLastChangedAt":"2024-07-09T14:06:16.585Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba5","user":{"_id":"6512aaceea5ccf1113e334fc","avatarUrl":"/avatars/a5b30f1c14cb2b2c633076719f824ec0.svg","isPro":false,"fullname":"Long Xing","user":"long-xing1","type":"user"},"name":"Long Xing","status":"claimed_verified","statusLastChangedAt":"2025-11-27T10:18:59.097Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba6","user":{"_id":"643e5d6a1d0e956d94bb3608","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/643e5d6a1d0e956d94bb3608/mTnu7Dts2RmDklHlp9Gqu.jpeg","isPro":false,"fullname":"Xiaoyang Wu","user":"Gofinge","type":"user"},"name":"Xiaoyang Wu","status":"admin_assigned","statusLastChangedAt":"2024-07-09T14:09:24.638Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba7","user":{"_id":"6310929fda0aaad2f70397a1","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6310929fda0aaad2f70397a1/YoHeYZezmCOXwL-qsiZOc.png","isPro":false,"fullname":"Tong Wu","user":"wutong16","type":"user"},"name":"Tong Wu","status":"claimed_verified","statusLastChangedAt":"2024-12-11T09:23:19.248Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba8","user":{"_id":"636317ed80c1a705a6eff396","avatarUrl":"/avatars/3db090e101b916d9256d0d3e043db71d.svg","isPro":false,"fullname":"Dahua Lin","user":"lindahua","type":"user"},"name":"Dahua Lin","status":"admin_assigned","statusLastChangedAt":"2024-07-09T14:08:43.006Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aeba9","user":{"_id":"65d5ec74cd05bc1eaa125040","avatarUrl":"/avatars/2de1b1539a86452c2c89570eeb02f5ab.svg","isPro":false,"fullname":"Xihui Liu","user":"XihuiLiu","type":"user"},"name":"Xihui Liu","status":"claimed_verified","statusLastChangedAt":"2024-10-16T09:44:52.854Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aebaa","user":{"_id":"64638c4d51fa6e63060521b5","avatarUrl":"/avatars/c863ace5b1dc788a341bcf4ddbdfaec1.svg","isPro":false,"fullname":"JIaqi","user":"Jiaqiwang","type":"user"},"name":"Jiaqi Wang","status":"admin_assigned","statusLastChangedAt":"2024-07-09T14:07:59.442Z","hidden":false},{"_id":"668cb6cc61b6eff5a82aebab","name":"Hengshuang Zhao","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/64b4eec4faa3181a5eab9c46/UABm3HnHCutmBzNcG5lox.mp4"],"publishedAt":"2024-07-08T17:59:55.000Z","submittedOnDailyAt":"2024-07-09T04:38:03.983Z","title":"Tailor3D: Customized 3D Assets Editing and Generation with Dual-Side\n Images","submittedOnDailyBy":{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","isPro":true,"fullname":"Jiaqi Wang","user":"myownskyW7","type":"user"},"summary":"Recent advances in 3D AIGC have shown promise in directly creating 3D objects\nfrom text and images, offering significant cost savings in animation and\nproduct design. However, detailed edit and customization of 3D assets remains a\nlong-standing challenge. Specifically, 3D Generation methods lack the ability\nto follow finely detailed instructions as precisely as their 2D image creation\ncounterparts. Imagine you can get a toy through 3D AIGC but with undesired\naccessories and dressing. To tackle this challenge, we propose a novel pipeline\ncalled Tailor3D, which swiftly creates customized 3D assets from editable\ndual-side images. We aim to emulate a tailor's ability to locally change\nobjects or perform overall style transfer. Unlike creating 3D assets from\nmultiple views, using dual-side images eliminates conflicts on overlapping\nareas that occur when editing individual views. Specifically, it begins by\nediting the front view, then generates the back view of the object through\nmulti-view diffusion. Afterward, it proceeds to edit the back views. Finally, a\nDual-sided LRM is proposed to seamlessly stitch together the front and back 3D\nfeatures, akin to a tailor sewing together the front and back of a garment. The\nDual-sided LRM rectifies imperfect consistencies between the front and back\nviews, enhancing editing capabilities and reducing memory burdens while\nseamlessly integrating them into a unified 3D representation with the LoRA\nTriplane Transformer. Experimental results demonstrate Tailor3D's effectiveness\nacross various 3D generation and editing tasks, including 3D generative fill\nand style transfer. It provides a user-friendly, efficient solution for editing\n3D assets, with each editing step taking only seconds to complete.","upvotes":13,"discussionId":"668cb6ce61b6eff5a82aec79","ai_summary":"Tailor3D addresses 3D asset customization by using dual-side images and a novel Dual-sided LRM for seamless integration, enhancing editing efficiency and consistency.","ai_keywords":["3D Generation","dual-side images","multi-view diffusion","Dual-sided LRM","LoRA Triplane Transformer","3D generative fill","style transfer"]},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","isPro":true,"fullname":"Jiaqi Wang","user":"myownskyW7","type":"user"},{"_id":"63dbbb5e06e5ca38798124ab","avatarUrl":"/avatars/b7598c94ed306788edb5ff6d9053a1e6.svg","isPro":false,"fullname":"M V","user":"MikrMrk","type":"user"},{"_id":"646ddc68deb963805b31b9f8","avatarUrl":"/avatars/2c266d28e947af7b447df426ab78423d.svg","isPro":false,"fullname":"martin","user":"martintomov","type":"user"},{"_id":"64f955c582673b2a07fbf0ad","avatarUrl":"/avatars/1c98c8be61f6580c1e4ee698fa5c0716.svg","isPro":false,"fullname":"hongyu","user":"learn12138","type":"user"},{"_id":"641b754d1911d3be6745cce9","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/641b754d1911d3be6745cce9/Ydjcjd4VuNUGj5Cd4QHdB.png","isPro":false,"fullname":"atayloraerospace","user":"Taylor658","type":"user"},{"_id":"63a75aba54f1d0225b082d44","avatarUrl":"/avatars/d11fc00b653bded5903bb936bd1fb55d.svg","isPro":false,"fullname":"jck","user":"jcka","type":"user"},{"_id":"651c240a37fecec1fe96c60b","avatarUrl":"/avatars/5af52af97b7907e138efecac0f20799b.svg","isPro":false,"fullname":"S.F.","user":"search-facility","type":"user"},{"_id":"6032802e1f993496bc14d9e3","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6032802e1f993496bc14d9e3/w6hr-DEQot4VVkoyRIBiy.png","isPro":false,"fullname":"Omar Sanseviero","user":"osanseviero","type":"user"},{"_id":"64316678dec2a70d8130aa9d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/EAS7OJwvyInle8J7IIBbw.jpeg","isPro":false,"fullname":"Levi Sverdlov","user":"Sverd","type":"user"},{"_id":"6370e6d53434f2764ce573ae","avatarUrl":"/avatars/6d1c9638a5bf427c6ea191db4cf40b9a.svg","isPro":false,"fullname":"Wuvin","user":"Wuvin","type":"user"},{"_id":"64de20c5808492ba6e65d124","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64de20c5808492ba6e65d124/58IX_TI5vJw73qS1knw56.jpeg","isPro":false,"fullname":"Zhang Mengchen","user":"Dubhe-zmc","type":"user"},{"_id":"663ccbff3a74a20189d4aa2e","avatarUrl":"/avatars/83a54455e0157480f65c498cd9057cf2.svg","isPro":false,"fullname":"Nguyen Van Thanh","user":"NguyenVanThanhHust","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":0}">
Tailor3D: Customized 3D Assets Editing and Generation with Dual-Side
Images
Abstract
Tailor3D addresses 3D asset customization by using dual-side images and a novel Dual-sided LRM for seamless integration, enhancing editing efficiency and consistency.
Recent advances in 3D AIGC have shown promise in directly creating 3D objects
from text and images, offering significant cost savings in animation and
product design. However, detailed edit and customization of 3D assets remains a
long-standing challenge. Specifically, 3D Generation methods lack the ability
to follow finely detailed instructions as precisely as their 2D image creation
counterparts. Imagine you can get a toy through 3D AIGC but with undesired
accessories and dressing. To tackle this challenge, we propose a novel pipeline
called Tailor3D, which swiftly creates customized 3D assets from editable
dual-side images. We aim to emulate a tailor's ability to locally change
objects or perform overall style transfer. Unlike creating 3D assets from
multiple views, using dual-side images eliminates conflicts on overlapping
areas that occur when editing individual views. Specifically, it begins by
editing the front view, then generates the back view of the object through
multi-view diffusion. Afterward, it proceeds to edit the back views. Finally, a
Dual-sided LRM is proposed to seamlessly stitch together the front and back 3D
features, akin to a tailor sewing together the front and back of a garment. The
Dual-sided LRM rectifies imperfect consistencies between the front and back
views, enhancing editing capabilities and reducing memory burdens while
seamlessly integrating them into a unified 3D representation with the LoRA
Triplane Transformer. Experimental results demonstrate Tailor3D's effectiveness
across various 3D generation and editing tasks, including 3D generative fill
and style transfer. It provides a user-friendly, efficient solution for editing
3D assets, with each editing step taking only seconds to complete.