HuGDiffusion: Generalizable Single-Image Human Rendering via 3D Gaussian Diffusion

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tang, Yingzhi, Zhang, Qijian, Hou, Junhui
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909848710414336
author Tang, Yingzhi
Zhang, Qijian
Hou, Junhui
author_facet Tang, Yingzhi
Zhang, Qijian
Hou, Junhui
contents We present HuGDiffusion, a generalizable 3D Gaussian splatting (3DGS) learning pipeline to achieve novel view synthesis (NVS) of human characters from single-view input images. Existing approaches typically require monocular videos or calibrated multi-view images as inputs, whose applicability could be weakened in real-world scenarios with arbitrary and/or unknown camera poses. In this paper, we aim to generate the set of 3DGS attributes via a diffusion-based framework conditioned on human priors extracted from a single image. Specifically, we begin with carefully integrated human-centric feature extraction procedures to deduce informative conditioning signals. Based on our empirical observations that jointly learning the whole 3DGS attributes is challenging to optimize, we design a multi-stage generation strategy to obtain different types of 3DGS attributes. To facilitate the training process, we investigate constructing proxy ground-truth 3D Gaussian attributes as high-quality attribute-level supervision signals. Through extensive experiments, our HuGDiffusion shows significant performance improvements over the state-of-the-art methods. Our code will be made publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2501_15008
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HuGDiffusion: Generalizable Single-Image Human Rendering via 3D Gaussian Diffusion
Tang, Yingzhi
Zhang, Qijian
Hou, Junhui
Computer Vision and Pattern Recognition
We present HuGDiffusion, a generalizable 3D Gaussian splatting (3DGS) learning pipeline to achieve novel view synthesis (NVS) of human characters from single-view input images. Existing approaches typically require monocular videos or calibrated multi-view images as inputs, whose applicability could be weakened in real-world scenarios with arbitrary and/or unknown camera poses. In this paper, we aim to generate the set of 3DGS attributes via a diffusion-based framework conditioned on human priors extracted from a single image. Specifically, we begin with carefully integrated human-centric feature extraction procedures to deduce informative conditioning signals. Based on our empirical observations that jointly learning the whole 3DGS attributes is challenging to optimize, we design a multi-stage generation strategy to obtain different types of 3DGS attributes. To facilitate the training process, we investigate constructing proxy ground-truth 3D Gaussian attributes as high-quality attribute-level supervision signals. Through extensive experiments, our HuGDiffusion shows significant performance improvements over the state-of-the-art methods. Our code will be made publicly available.
title HuGDiffusion: Generalizable Single-Image Human Rendering via 3D Gaussian Diffusion
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2501.15008