Home

Awesome

<h3 align="center">Modelling complex vector drawings with stroke-clouds</h3> <p align="center"> ICLR 2024 Poster Paper <br /> <br /> <br /> <a href="https://iclr.cc/virtual/2024/poster/18757">Paper</a> · <a href="https://drive.google.com/file/d/1rv8MGfiAv6lwddGiSLTCqK2Mfltqnatk/view?usp=sharing">Data</a> · <a href="https://drive.google.com/drive/folders/1e61EzE33T7foYsLqr-B3IGHLtgVX2MPD?usp=sharing">Models</a> </p> </div> <!-- ABOUT THE PROJECT -->

<img src="https://github.com/Co-do/Stroke-Cloud/assets/123647750/411131f4-4826-4763-a485-69cd929a8e26" width="200" height="200"> <img src="https://github.com/Co-do/Stroke-Cloud/assets/123647750/ea4ce9be-d05c-4393-9ed8-91152aff3c12" width="200" height="200"> <img src="https://github.com/Co-do/Stroke-Cloud/assets/123647750/4f554f6f-0f9b-464e-9bc8-6faffe1392e0" width="200" height="200"> <img src="https://github.com/Co-do/Stroke-Cloud/assets/123647750/4172389c-a0b4-4ecd-866f-296790c0706e" width="200" height="200">

<!-- GETTING STARTED -->

Introduction

This is the offical repository for "Modelling complex vector drawings with stroke-clouds". The data set and pre-trained models can be downloaded from the links given. Instructions for inference and training both the srm and lsg are given.

Environment

lxml - 4.9

lightning - 2.0.9

pytorch-lightning - 2.0.3

torch - 2.2.0

Training

SRM:

  1. Add your wandb key or change the logger in srm_train.py.
  2. Download the training data.
  3. Run srm_train.py

LSG:

  1. Generate latent codes with srm_test.py
  2. Add your wandb key or change the logger in lsg_train.py.
  3. Run lsg_train.py

Inference

  1. Download the models and run lsg_test.py
  2. Or run lsg_test.py with your own trained models.
<!-- LICENSE -->

License

Distributed under the MIT License. See LICENSE.txt for more information.

<!-- CONTACT -->

Contact

Alexander Ashcroft - aa05377@surrey.ac.uk