This repository contains code and study materials for the paper:
Linping Yuan, Ziqi Zhou, Jian Zhao, Yiqiu Guo, Fan Du, Huamin Qu. InfoColorizer: Interactive Recommendation of Color Palettes for Infographics. IEEE Transactions on Visualization and Computer Graphics, 2021.
An ArXiv copy can also be downloaded here.
InfoColorizer is a tool that allows users to effectively obtain high-quality color palettes during infographics creation. It consists of a recommendation engine and a visual interface. From a large infographic dataset, we extract a set of features that embeds both structure and color information of infographics (the red arrow). We then train a deep learning model, VAEAC, that characterizes good color design practices in the data, to construct our recommendation engine. With the visual interface, users can obtain recommended color palettes, specify various color preferences and constraints, preview and edit infographics, and retrieve new recommendations in an iterative manner (the blue arrows).
As described in Section 5.1.1, we extract features at different levels and construct a tree that captures spatial relationship of infographic elements. The corresponding code is mainly at ./backend/treeconstructor.
- cd frontend
- npm install
- npm run serve
- cd backend
- pip install -r requirements.txt (suggest using virtual environment https://docs.python.org/3/tutorial/venv.html)
- bash run-data-backend.sh
@article{yuan2021infocolorizer,
title={InfoColorizer: Interactive Recommendation of Color Palettes for Infographics},
author={Yuan, Lin-Ping and Zhou, Ziqi and Zhao, Jian and Guo, Yiqiu and Du, Fan and Qu, Huamin},
journal={IEEE Transactions on Visualization and Computer Graphics},
year={2021}
}