This is a fork of the original source code for the paper "ArtCoder: An End-to-end Method for Generating Scanning-robust Stylized QR Codes".
There has been many changes and improvements, including:
- A complete web interface to select images and change the training parameters
- Optimizations (I don't recall the numbers exactly but it's at least a x2 speedup)
- Automatic segmentation of the input data for minimal encoding size
- Base Qr code generation
- Qart code implementation so the base code fits the target image, using saliency and edge detection
Style&Content | Output |
- Linux or Windows
- CPU or NVIDIA GPU + CUDA CuDNN
- Python 3
- Pytorch 1.7.0
- Other Python libraries I forgot about but you'll find out as you try to run it
- Clone this repo:
git clone https://github.com/SwordHolderSH/ArtCoder
cd ArtCoder
- Install PyTorch and other dependencies.
- Start the server
python ./src/app/main.py
It will start a webserver you can visit at http://localhost:8056 (and a webserver at ws://localhost:8057)