Skip to content
/ Adain Public

A implementation of Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization

Notifications You must be signed in to change notification settings

Klayand/Adain

Repository files navigation

✨ Adain ✨

@Author: Zikai Zhou

Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization(unofficial implementation) 🚀

Results

Overview

Using pretraining weights of VGG, adain can do image transfer tasks in real time!

Technical Features

  • Small Computational Resources: The size of model weight file is only about 30MB.

  • Real Time Inference: Inference with model in real time.

  • Multi-style Fusion: Support multi-style transfer. Hosi-Sketch

  • Color Preserving: Support preserving the color of original image with transferred image style. self-portrait

Getting Started

  1. Clone the repository:

    git clone https://github.com/Klayand/Adain.git
  2. Navigate to the project directory:

    cd Adain
  3. Install the dependencies:

    pip install -r requirements.txt
    

Usage

  1. Train model:

    python train.py --user-envs
    
  2. Inference model: python test.py --user-envs

    
    

About

A implementation of Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages