This repository contains the source code for AddCR: A Data-Driven Cartoon Remastering
This code has been tested with Pytorch 1.8.1 and Cuda 11.1
conda create -n addcr python==3.8.3
conda activate addcr
pip install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio==0.8.1 -f https://download.pytorch.org/whl/torch_stable.html
pip install basicsr opencv-python numpy
python setup.py develop
You can demo a trained model on a sequence of frames
python test.py --cuda_num 0 --input /your-own-path/inputs/ --output /your-own-path/results/
--cuda_num: the number of your cuda
--input: the path on which you stored the unprocessed frames
--output: the path of final output
RealESRGAN: the code is modified from https://github.com/xinntao/Real-ESRGAN