DiDE:Direct Injection with Color-Texture DEcoupling for 3D Stylization

2026-10-01

AI Summary

This study addresses the challenge of independently controlling color and texture in existing 3D stylization approaches by proposing DiDE, the first training-free decoupled framework. Leveraging the overcompleteness of latent spaces, this work introduces a novel channel partitioning mechanism that enables interference-free injection of dual reference signals within self-attention layers. Furthermore, DiDE incorporates rectified flow models to facilitate efficient multi-reference image injection. Experimental results demonstrate that DiDE significantly outperforms existing baseline methods on the Disen3D-Bench benchmark, substantially improving both the fidelity and the decoupled controllability of color and texture in 3D stylization tasks.

Abstract

Recent advances in rectified flow-based image-to-3D generative models have enabled high-fidelity 3D asset generation. Building on this, a growing line of work has exploited these strong 3D priors for training-free stylization, transferring visual attributes from a reference image onto a generated 3D asset. However, existing methods enforce an all-or-nothing paradigm: color and texture are transferred jointly, with no mechanism to control them independently -- a limitation we formalize as Disentangled 3D Stylization(Disen3D). To address this, we propose DiDE, the first training-free framework for Disen3D. Key to our approach is the observation that the structured latent space of image-to-3D models is overcomplete with respect to texture: texture information occupies only a small subset of the style-significant channels, leaving a free subspace available for independent color encoding. DiDE exploits this via a channel partition mechanism that processes a content image, a texture reference, and a color reference through dedicated branches and composes both style signals interference-free at every self-attention layer, preserving content geometry throughout. Experiments on Disen3D-Bench, our newly collected multi-reference benchmark, show that DiDE consistently outperforms 2D and 3D stylization baselines in color fidelity, texture transfer, and content preservation.

Problem

• 3D Stylization
• Disentangled Stylization
• Color-Texture Decoupling
• Image-to-3D Generation
• Training-free

Innovation

• Disentangled 3D Stylization
• Training-free
• Color-Texture Decoupling
• Channel Partition Mechanism
• Rectified Flow