4DStyleGaussian: Zero-shot 4D Style Transfer with Gaussian Splatting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liang, Wanlin, Xu, Hongbin, Chen, Weitao, Xiao, Feng, Kang, Wenxiong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910648961597440
author Liang, Wanlin
Xu, Hongbin
Chen, Weitao
Xiao, Feng
Kang, Wenxiong
author_facet Liang, Wanlin
Xu, Hongbin
Chen, Weitao
Xiao, Feng
Kang, Wenxiong
contents 3D neural style transfer has gained significant attention for its potential to provide user-friendly stylization with spatial consistency. However, existing 3D style transfer methods often fall short in terms of inference efficiency, generalization ability, and struggle to handle dynamic scenes with temporal consistency. In this paper, we introduce 4DStyleGaussian, a novel 4D style transfer framework designed to achieve real-time stylization of arbitrary style references while maintaining reasonable content affinity, multi-view consistency, and temporal coherence. Our approach leverages an embedded 4D Gaussian Splatting technique, which is trained using a reversible neural network for reducing content loss in the feature distillation process. Utilizing the 4D embedded Gaussians, we predict a 4D style transformation matrix that facilitates spatially and temporally consistent style transfer with Gaussian Splatting. Experiments demonstrate that our method can achieve high-quality and zero-shot stylization for 4D scenarios with enhanced efficiency and spatial-temporal consistency.
format Preprint
id arxiv_https___arxiv_org_abs_2410_10412
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle 4DStyleGaussian: Zero-shot 4D Style Transfer with Gaussian Splatting
Liang, Wanlin
Xu, Hongbin
Chen, Weitao
Xiao, Feng
Kang, Wenxiong
Computer Vision and Pattern Recognition
3D neural style transfer has gained significant attention for its potential to provide user-friendly stylization with spatial consistency. However, existing 3D style transfer methods often fall short in terms of inference efficiency, generalization ability, and struggle to handle dynamic scenes with temporal consistency. In this paper, we introduce 4DStyleGaussian, a novel 4D style transfer framework designed to achieve real-time stylization of arbitrary style references while maintaining reasonable content affinity, multi-view consistency, and temporal coherence. Our approach leverages an embedded 4D Gaussian Splatting technique, which is trained using a reversible neural network for reducing content loss in the feature distillation process. Utilizing the 4D embedded Gaussians, we predict a 4D style transformation matrix that facilitates spatially and temporally consistent style transfer with Gaussian Splatting. Experiments demonstrate that our method can achieve high-quality and zero-shot stylization for 4D scenarios with enhanced efficiency and spatial-temporal consistency.
title 4DStyleGaussian: Zero-shot 4D Style Transfer with Gaussian Splatting
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.10412