Saved in:
Bibliographic Details
Main Authors: Yang, Xianghui, Shi, Huiwen, Zhang, Bowen, Yang, Fan, Wang, Jiacheng, Zhao, Hongxu, Liu, Xinhai, Wang, Xinzhou, Lin, Qingxiang, Yu, Jiaao, Wang, Lifu, Xu, Jing, He, Zebin, Chen, Zhuo, Liu, Sicong, Wu, Junta, Lian, Yihang, Yang, Shaoxiong, Liu, Yuhong, Yang, Yong, Wang, Di, Jiang, Jie, Guo, Chunchao
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.02293
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912200508047360
author Yang, Xianghui
Shi, Huiwen
Zhang, Bowen
Yang, Fan
Wang, Jiacheng
Zhao, Hongxu
Liu, Xinhai
Wang, Xinzhou
Lin, Qingxiang
Yu, Jiaao
Wang, Lifu
Xu, Jing
He, Zebin
Chen, Zhuo
Liu, Sicong
Wu, Junta
Lian, Yihang
Yang, Shaoxiong
Liu, Yuhong
Yang, Yong
Wang, Di
Jiang, Jie
Guo, Chunchao
author_facet Yang, Xianghui
Shi, Huiwen
Zhang, Bowen
Yang, Fan
Wang, Jiacheng
Zhao, Hongxu
Liu, Xinhai
Wang, Xinzhou
Lin, Qingxiang
Yu, Jiaao
Wang, Lifu
Xu, Jing
He, Zebin
Chen, Zhuo
Liu, Sicong
Wu, Junta
Lian, Yihang
Yang, Shaoxiong
Liu, Yuhong
Yang, Yong
Wang, Di
Jiang, Jie
Guo, Chunchao
contents While 3D generative models have greatly improved artists' workflows, the existing diffusion models for 3D generation suffer from slow generation and poor generalization. To address this issue, we propose a two-stage approach named Hunyuan3D 1.0 including a lite version and a standard version, that both support text- and image-conditioned generation. In the first stage, we employ a multi-view diffusion model that efficiently generates multi-view RGB in approximately 4 seconds. These multi-view images capture rich details of the 3D asset from different viewpoints, relaxing the tasks from single-view to multi-view reconstruction. In the second stage, we introduce a feed-forward reconstruction model that rapidly and faithfully reconstructs the 3D asset given the generated multi-view images in approximately 7 seconds. The reconstruction network learns to handle noises and in-consistency introduced by the multi-view diffusion and leverages the available information from the condition image to efficiently recover the 3D structure. Our framework involves the text-to-image model, i.e., Hunyuan-DiT, making it a unified framework to support both text- and image-conditioned 3D generation. Our standard version has 3x more parameters than our lite and other existing model. Our Hunyuan3D 1.0 achieves an impressive balance between speed and quality, significantly reducing generation time while maintaining the quality and diversity of the produced assets.
format Preprint
id arxiv_https___arxiv_org_abs_2411_02293
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Hunyuan3D 1.0: A Unified Framework for Text-to-3D and Image-to-3D Generation
Yang, Xianghui
Shi, Huiwen
Zhang, Bowen
Yang, Fan
Wang, Jiacheng
Zhao, Hongxu
Liu, Xinhai
Wang, Xinzhou
Lin, Qingxiang
Yu, Jiaao
Wang, Lifu
Xu, Jing
He, Zebin
Chen, Zhuo
Liu, Sicong
Wu, Junta
Lian, Yihang
Yang, Shaoxiong
Liu, Yuhong
Yang, Yong
Wang, Di
Jiang, Jie
Guo, Chunchao
Computer Vision and Pattern Recognition
Artificial Intelligence
While 3D generative models have greatly improved artists' workflows, the existing diffusion models for 3D generation suffer from slow generation and poor generalization. To address this issue, we propose a two-stage approach named Hunyuan3D 1.0 including a lite version and a standard version, that both support text- and image-conditioned generation. In the first stage, we employ a multi-view diffusion model that efficiently generates multi-view RGB in approximately 4 seconds. These multi-view images capture rich details of the 3D asset from different viewpoints, relaxing the tasks from single-view to multi-view reconstruction. In the second stage, we introduce a feed-forward reconstruction model that rapidly and faithfully reconstructs the 3D asset given the generated multi-view images in approximately 7 seconds. The reconstruction network learns to handle noises and in-consistency introduced by the multi-view diffusion and leverages the available information from the condition image to efficiently recover the 3D structure. Our framework involves the text-to-image model, i.e., Hunyuan-DiT, making it a unified framework to support both text- and image-conditioned 3D generation. Our standard version has 3x more parameters than our lite and other existing model. Our Hunyuan3D 1.0 achieves an impressive balance between speed and quality, significantly reducing generation time while maintaining the quality and diversity of the produced assets.
title Hunyuan3D 1.0: A Unified Framework for Text-to-3D and Image-to-3D Generation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2411.02293