🥳Awesome Style Transfer with Diffusion Models

🤗 Introduction

• This repository contains a curated list of Style Transfer with Diffusion Models. The resources are organized into two main categories: Image Synthesis and Video Synthesis.

• We welcome any contributions and suggestions to our repository or the addition of your own work. Feel free to make a pull request or leave your comments!!

📋 Contents

• 🤗 Introduction
• 📋 Contents
• 🤔 Perspectives
• 📍 Image Synthesis
  • Fundation & Application Models
  • Image-Encoder Based
  • Swap Features In Attention Layer
  • Optimization-Based Methods
  • LoRA-Based
• 📍 Video Synthesis
• 📍 3D Generation
• 📍 Style Dataset
• 📍 Related Research and Surveys

🤔 Perspectives

A personal view from the maintainers, September 2026.

Style is vague and hard to define. Seen from one angle, style is also a kind of semantics, and the difficulty of this task lies in separating out the many different kinds of semantics. The diversity of style itself then makes that separation and disentanglement harder.

Style transfer is also more of an application, and every user has their own view of it.

We reorganized the paper taxonomy to focus more on how a method encodes the style image. For example, an image encoder is a feed-forward any-style method, while a LoRA-based method applies to a single style.

📍 Image Synthesis

In the context of Image Style Transfer, the techniques can be categorized into two primary types: text-driven style transfer and image-driven style transfer and the techniques are organized by the mechanism that carries the style: image encoders and adapters, attention feature swapping, LoRA, and optimization-based methods.

Foundation & Application Models

Image-Encoder Based

Swap Features In Attention Layer

Optimization-Based Methods

LoRA-Based Methods

📍 Video Synthesis

📍 3D Generation

📍 Style Dataset

📍 Related Research and Surveys