Capacity-Controlled Multi-View Stylization of 3D Gaussian Splatting

Abstract

While 3D Gaussian Splatting (3DGS) provides an efficient and explicit representation for novel view synthesis, enforcing stylistic coherence across viewpoints remains challenging. Existing 3D stylization methods typically apply 2D feature-matching losses independently per rendered view, which leads to unstable style allocation, many-to-one feature reuse, and limited cross-view consistency. We propose a capacity-controlled framework for multi-view stylization of 3DGS, grounded in optimal transport. Specifically, we reformulate local style matching as a semi-balanced optimal transport problem. By introducing explicit column-capacity constraints with tunable strength, our formulation mitigates many-to-one matching and enables controllable allocation of style features. This transport-based objective provides a principled mechanism for balancing feature coverage and stylistic diversity while maintaining stable correspondences across viewpoints. To further enhance cross-view coherence, we incorporate a novel cross-view matching guidance to constrain correspondences between scene content and style patterns. In addition, we introduce several geometric regularizations to enhance the vanilla 3DGS, thereby enabling optimized Gaussian primitives to represent finer-grained textures during stylization. Extensive experiments demonstrate that our approach significantly improves multi-view stylistic consistency and produces stable, expressive 3D stylizations while preserving the core semantic structure of the scene.

Method

Our framework contains two stages: enhanced reconstruction and capacity-controlled stylization. In the reconstruction stage, we optimize the 3D Gaussians using estimated depth as geometric constraints to achieve more accurate scene reconstruction. Meanwhile, we regularize the scale and shape of the primitives to better represent stylized textures later. Then, in the stylization stage, we introduce a Capacity-Controlled Feature Transport (CCFT) loss and Cross-View Matching Guidance to obtain coherent stylized views. With a content loss that preserves scene structure, we update the color parameters of 3D Gaussians, yielding the final stylized 3D scene.

Capacity-Controlled Feature Transport

The capacity of each style feature is represented by the size of its node, with larger nodes indicating greater capacity. (a) When there's no capacity constraint, multiple rendering features are mapped to the same style feature, and matching degrades to nearest-neighbor search, resulting in the many-to-one issue. (b) With capacity control, the matching of each style feature tends to be equalized, and more style features are utilized.

Cross-view Matching Guidance

Inspired by Guided Correspondence, We introduce a guidance map to improve matching coherence across viewpoints by reaggregating rendered features according to the transport matrix. The element of the guidance map indicates which rendered features are matched to the corresponding style feature at viewpoint v − 1. This guidance encourages similar rendered features across different viewpoints to match the same corresponding style feature.