Three-Dimensional Gaussian Style Transfer Method Based on Two-Dimensional Priors and Iterative Optimization
Open Access
- 3 September 2025
- journal article
- research article
- Published by MDPI AG in Applied Sciences
- Vol. 15 (17) , 9678
Abstract
To address the limitations of existing optimization-based 3D style transfer methods in terms of visual quality, 3D consistency, and real-time rendering performance, we propose a novel 3D Gaussian scene style transfer method based on 2D priors and iterative optimization. Our approach introduces a progressive training pipeline that alternates between fine-tuning the 3D Gaussian field and updating a set of supervised stylized images. By gradually injecting style information into the 3D scene through iterative refinement, the method effectively preserves the geometric structure and spatial coherence across viewpoints. Furthermore, we incorporated a pre-trained stable diffusion model as a 2D prior to guide the style adaptation of the 3D Gaussian representation. The combination of diffusion priors and differentiable 3D Gaussian rendering enables high-fidelity style transfer while maintaining real-time rendering capability. Extensive experiments demonstrate that our method significantly improves the visual quality and multi-view consistency of 3D stylized scenes, offering an effective and efficient solution for real-time 3D scene stylization.Keywords
Funding Information
- Zhejiang Province Program (2025C01068, 2024C03263, LZ25F020006, 2024009, 62172365, 2025Z052, 2025Z062, 2022Z167, 2023Z137)
- Zhejiang Provincial Cultural Relics Protection Science and Technology Project (2025C01068, 2024C03263, LZ25F020006, 2024009, 62172365, 2025Z052, 2025Z062, 2022Z167, 2023Z137)
- National Program of China (2025C01068, 2024C03263, LZ25F020006, 2024009, 62172365, 2025Z052, 2025Z062, 2022Z167, 2023Z137)
- Macau project: Key technology research and display system development for new personalized controllable dressing dynamic display, and Ningbo Science and Technology Plan Project (2025C01068, 2024C03263, LZ25F020006, 2024009, 62172365, 2025Z052, 2025Z062, 2022Z167, 2023Z137)
This publication has 22 references indexed in Scilit:
- StyleSplat: 3D Object Style Transfer with Gaussian SplattingarXiv, 2024
- ARF: Artistic Radiance Fields2022
- NeRFCommunications of the ACM, 2021
- Local light field fusionACM Transactions on Graphics, 2019
- Arbitrary Style Transfer With Style-Attentional Networks2019
- Tanks and templesACM Transactions on Graphics, 2017
- Image Style Transfer Using Convolutional Neural Networks2016
- Perceptual Losses for Real-Time Style Transfer and Super-Resolution2016
- Pixelwise View Selection for Unstructured Multi-View Stereo2016
- Image quality assessment: from error visibility to structural similarityIEEE Transactions on Image Processing, 2004