AnyStyle: Single-Pass Multimodal Stylization for 3D Gaussian Splatting

2026-02-03

AI Summary

Existing feedforward 3D reconstruction methods that operate without camera pose priors lack flexible and controllable stylization capabilities. This work proposes AnyStyle, the first zero-shot multimodal stylization framework that supports both text- and image-guided appearance control, enabling simultaneous 3D Gaussian splatting reconstruction and style manipulation in a single feedforward pass. Built upon a lightweight modular architecture, AnyStyle achieves high geometric fidelity while offering strong stylistic controllability—all without requiring pose priors. Experimental results demonstrate that AnyStyle significantly outperforms existing feedforward stylization approaches in terms of style quality while preserving accurate geometry, and user studies further confirm its superior stylization performance.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Natural Language Processing: Sentiment Analysis, Stylistic Analysis, and Argument Mining

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

The growing demand for rapid and scalable 3D asset creation has driven interest in feed-forward 3D reconstruction methods, with 3D Gaussian Splatting (3DGS) emerging as an effective scene representation. While recent approaches have demonstrated pose-free reconstruction from unposed image collections, integrating stylization or appearance control into such pipelines remains underexplored. Existing attempts largely rely on image-based conditioning, which limits both controllability and flexibility. In this work, we introduce AnyStyle, a feed-forward 3D reconstruction and stylization framework that enables pose-free, zero-shot stylization through multimodal conditioning. Our method supports both textual and visual style inputs, allowing users to control the scene appearance using natural language descriptions or reference images. We propose a modular stylization architecture that requires only minimal architectural modifications and can be integrated into existing feed-forward 3D reconstruction backbones. Experiments demonstrate that AnyStyle improves style controllability over prior feed-forward stylization methods while preserving high-quality geometric reconstruction. A user study further confirms that AnyStyle achieves superior stylization quality compared to an existing state-of-the-art approach. Repository: https://github.com/joaxkal/AnyStyle.

Problem

• 3D Gaussian Splatting
• stylization
• pose-free reconstruction
• multimodal conditioning
• appearance control

Innovation

• 3D Gaussian Splatting
• multimodal stylization
• pose-free reconstruction
• zero-shot stylization
• feed-forward 3D reconstruction