Multi-level Dynamic Style Transfer for NeRFs

2025-10-01

AI Summary

Existing NeRF-based style transfer methods struggle to balance content fidelity and artistic expressiveness. To address this, we propose MDS-NeRF, a Multi-level Dynamic Style transfer framework for 3D scenes. Our method introduces hierarchical multi-scale feature grids to explicitly model spatial structure at varying levels of detail; designs a dynamic style injection module that adaptively fuses style features across scales; and employs a cascaded decoder architecture to jointly optimize the content radiance field and stylized rendering. Crucially, MDS-NeRF enables view-consistent stylization of novel viewpoints using only a single 3D style reference. Experiments demonstrate that our approach preserves multi-scale geometric and semantic structure while significantly improving both stylization quality and content fidelity, outperforming state-of-the-art methods on multiple benchmarks.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Natural Language Processing: Sentiment Analysis, Stylistic Analysis, and Argument Mining

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• 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

As the application of neural radiance fields (NeRFs) in various 3D vision tasks continues to expand, numerous NeRF-based style transfer techniques have been developed. However, existing methods typically integrate style statistics into the original NeRF pipeline, often leading to suboptimal results in both content preservation and artistic stylization. In this paper, we present multi-level dynamic style transfer for NeRFs (MDS-NeRF), a novel approach that reengineers the NeRF pipeline specifically for stylization and incorporates an innovative dynamic style injection module. Particularly, we propose a multi-level feature adaptor that helps generate a multi-level feature grid representation from the content radiance field, effectively capturing the multi-scale spatial structure of the scene. In addition, we present a dynamic style injection module that learns to extract relevant style features and adaptively integrates them into the content patterns. The stylized multi-level features are then transformed into the final stylized view through our proposed multi-level cascade decoder. Furthermore, we extend our 3D style transfer method to support omni-view style transfer using 3D style references. Extensive experiments demonstrate that MDS-NeRF achieves outstanding performance for 3D style transfer, preserving multi-scale spatial structures while effectively transferring stylistic characteristics.

Problem

• Improving 3D style transfer quality in neural radiance fields
• Preserving multi-scale spatial structures during stylization
• Adaptively integrating style features into content patterns

Innovation

• Multi-level feature adaptor captures scene spatial structures
• Dynamic style injection module integrates style features adaptively
• Multi-level cascade decoder transforms features into stylized views