IDOL: Instant Photorealistic 3D Human Creation from a Single Image

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhuang, Yiyu, Lv, Jiaxi, Wen, Hao, Shuai, Qing, Zeng, Ailing, Zhu, Hao, Chen, Shifeng, Yang, Yujiu, Cao, Xun, Liu, Wei
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916661543567360
author Zhuang, Yiyu
Lv, Jiaxi
Wen, Hao
Shuai, Qing
Zeng, Ailing
Zhu, Hao
Chen, Shifeng
Yang, Yujiu
Cao, Xun
Liu, Wei
author_facet Zhuang, Yiyu
Lv, Jiaxi
Wen, Hao
Shuai, Qing
Zeng, Ailing
Zhu, Hao
Chen, Shifeng
Yang, Yujiu
Cao, Xun
Liu, Wei
contents Creating a high-fidelity, animatable 3D full-body avatar from a single image is a challenging task due to the diverse appearance and poses of humans and the limited availability of high-quality training data. To achieve fast and high-quality human reconstruction, this work rethinks the task from the perspectives of dataset, model, and representation. First, we introduce a large-scale HUman-centric GEnerated dataset, HuGe100K, consisting of 100K diverse, photorealistic sets of human images. Each set contains 24-view frames in specific human poses, generated using a pose-controllable image-to-multi-view model. Next, leveraging the diversity in views, poses, and appearances within HuGe100K, we develop a scalable feed-forward transformer model to predict a 3D human Gaussian representation in a uniform space from a given human image. This model is trained to disentangle human pose, body shape, clothing geometry, and texture. The estimated Gaussians can be animated without post-processing. We conduct comprehensive experiments to validate the effectiveness of the proposed dataset and method. Our model demonstrates the ability to efficiently reconstruct photorealistic humans at 1K resolution from a single input image using a single GPU instantly. Additionally, it seamlessly supports various applications, as well as shape and texture editing tasks. Project page: https://yiyuzhuang.github.io/IDOL/.
format Preprint
id arxiv_https___arxiv_org_abs_2412_14963
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle IDOL: Instant Photorealistic 3D Human Creation from a Single Image
Zhuang, Yiyu
Lv, Jiaxi
Wen, Hao
Shuai, Qing
Zeng, Ailing
Zhu, Hao
Chen, Shifeng
Yang, Yujiu
Cao, Xun
Liu, Wei
Computer Vision and Pattern Recognition
Graphics
Machine Learning
68U05, 68T07, 68T45
I.3.7; I.2.10; I.2.6
Creating a high-fidelity, animatable 3D full-body avatar from a single image is a challenging task due to the diverse appearance and poses of humans and the limited availability of high-quality training data. To achieve fast and high-quality human reconstruction, this work rethinks the task from the perspectives of dataset, model, and representation. First, we introduce a large-scale HUman-centric GEnerated dataset, HuGe100K, consisting of 100K diverse, photorealistic sets of human images. Each set contains 24-view frames in specific human poses, generated using a pose-controllable image-to-multi-view model. Next, leveraging the diversity in views, poses, and appearances within HuGe100K, we develop a scalable feed-forward transformer model to predict a 3D human Gaussian representation in a uniform space from a given human image. This model is trained to disentangle human pose, body shape, clothing geometry, and texture. The estimated Gaussians can be animated without post-processing. We conduct comprehensive experiments to validate the effectiveness of the proposed dataset and method. Our model demonstrates the ability to efficiently reconstruct photorealistic humans at 1K resolution from a single input image using a single GPU instantly. Additionally, it seamlessly supports various applications, as well as shape and texture editing tasks. Project page: https://yiyuzhuang.github.io/IDOL/.
title IDOL: Instant Photorealistic 3D Human Creation from a Single Image
topic Computer Vision and Pattern Recognition
Graphics
Machine Learning
68U05, 68T07, 68T45
I.3.7; I.2.10; I.2.6
url https://arxiv.org/abs/2412.14963