Saved in:
Bibliographic Details
Main Authors: Sun, Jingxiang, Peng, Cheng, Shao, Ruizhi, Guo, Yuan-Chen, Zhao, Xiaochen, Li, Yangguang, Cao, Yanpei, Zhang, Bo, Liu, Yebin
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.12928
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913550185791488
author Sun, Jingxiang
Peng, Cheng
Shao, Ruizhi
Guo, Yuan-Chen
Zhao, Xiaochen
Li, Yangguang
Cao, Yanpei
Zhang, Bo
Liu, Yebin
author_facet Sun, Jingxiang
Peng, Cheng
Shao, Ruizhi
Guo, Yuan-Chen
Zhao, Xiaochen
Li, Yangguang
Cao, Yanpei
Zhang, Bo
Liu, Yebin
contents We introduce DreamCraft3D++, an extension of DreamCraft3D that enables efficient high-quality generation of complex 3D assets. DreamCraft3D++ inherits the multi-stage generation process of DreamCraft3D, but replaces the time-consuming geometry sculpting optimization with a feed-forward multi-plane based reconstruction model, speeding up the process by 1000x. For texture refinement, we propose a training-free IP-Adapter module that is conditioned on the enhanced multi-view images to enhance texture and geometry consistency, providing a 4x faster alternative to DreamCraft3D's DreamBooth fine-tuning. Experiments on diverse datasets demonstrate DreamCraft3D++'s ability to generate creative 3D assets with intricate geometry and realistic 360° textures, outperforming state-of-the-art image-to-3D methods in quality and speed. The full implementation will be open-sourced to enable new possibilities in 3D content creation.
format Preprint
id arxiv_https___arxiv_org_abs_2410_12928
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DreamCraft3D++: Efficient Hierarchical 3D Generation with Multi-Plane Reconstruction Model
Sun, Jingxiang
Peng, Cheng
Shao, Ruizhi
Guo, Yuan-Chen
Zhao, Xiaochen
Li, Yangguang
Cao, Yanpei
Zhang, Bo
Liu, Yebin
Computer Vision and Pattern Recognition
We introduce DreamCraft3D++, an extension of DreamCraft3D that enables efficient high-quality generation of complex 3D assets. DreamCraft3D++ inherits the multi-stage generation process of DreamCraft3D, but replaces the time-consuming geometry sculpting optimization with a feed-forward multi-plane based reconstruction model, speeding up the process by 1000x. For texture refinement, we propose a training-free IP-Adapter module that is conditioned on the enhanced multi-view images to enhance texture and geometry consistency, providing a 4x faster alternative to DreamCraft3D's DreamBooth fine-tuning. Experiments on diverse datasets demonstrate DreamCraft3D++'s ability to generate creative 3D assets with intricate geometry and realistic 360° textures, outperforming state-of-the-art image-to-3D methods in quality and speed. The full implementation will be open-sourced to enable new possibilities in 3D content creation.
title DreamCraft3D++: Efficient Hierarchical 3D Generation with Multi-Plane Reconstruction Model
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.12928