Mixture of Style Experts for Diverse Image Stylization
This work addresses the need for more complex and semantic-aware image stylization for applications in digital art and media, representing an incremental improvement over prior diffusion-based methods.
The paper tackles the problem of limited color-driven transformations in diffusion-based image stylization by introducing StyleExpert, a semantic-aware framework that uses a Mixture of Experts to handle diverse styles from textures to deep semantics, resulting in outperformance over existing methods in preserving semantics and material details with generalization to unseen styles.
Diffusion-based stylization has advanced significantly, yet existing methods are limited to color-driven transformations, neglecting complex semantics and material details.We introduce StyleExpert, a semantic-aware framework based on the Mixture of Experts (MoE). Our framework employs a unified style encoder, trained on our large-scale dataset of content-style-stylized triplets, to embed diverse styles into a consistent latent space. This embedding is then used to condition a similarity-aware gating mechanism, which dynamically routes styles to specialized experts within the MoE architecture. Leveraging this MoE architecture, our method adeptly handles diverse styles spanning multiple semantic levels, from shallow textures to deep semantics. Extensive experiments show that StyleExpert outperforms existing approaches in preserving semantics and material details, while generalizing to unseen styles. Our code and collected images are available at the project page: https://hh-lg.github.io/StyleExpert-Page/.