Pluggable Style Representation Learning for Multi-Style Transfer (ACCV 2024)

Pytorch implementation of paper Pluggable Style Representation Learning for Multi-Style Transfer (ACCV 2024 paper).

Visual examples:

Installation

• Install python 3.8.0, torch 2.0.0, CUDA 11.7 and other essential packages (Note that using other versions of packages may affect performance).
• Clone this repo

git clone https://github.com/Chernobyllight/SaMST
cd SaMST

Test

• Get into evaluation codes folder './test_model/test1/test.yml'

  cd ./test_model/test/
  

• We provide pretrained models trained on 10 styles in './checkpoints/'. Before testing, specify the model checkpoint in 'test.yml'.

  model: './checkpoints/epoch_20.model'
  

  Or change the path with your own absolute path. Then you need to specify the number of styles to train the whole model.

  style_num: 10
  

  Finally, specify your own test content image folder:

  content_image_dir: './content/'
  

• Run 'test.py'

  python test.py
  

Train

We provide two training pipelines. If you have a large number of styles, please get into train1. In contrast, if you just train the model on several styles, you can get into train2 to strike a fast convergence. There is no difference on the two training pipelines' option setting. Here is a example of pipeline train1.

• Get into training codes folder './train_model/train1/train.yml':

  cd ./train_model/train1/train.yml
  

• Specify the dataset paths in 'train.yml':

  dataset: your_training_content_path
  style_image: your_training_style_path
  

  Or change the path with your own dataset absolute path. Then specify the checkpoint saving path:

  save_model_dir: './checkpoint/'
  

• Run 'train.py'

  python train.py