TCI_Cycle-free CycleGAN

Paper

Cycle-free CycleGAN using Invertible Generator for Unsupervised Low-Dose CT Denoising (IEEE TCI, T. Kwon et al.)

Sample Data

Public dataset that we used was from Low Dose CT Grand Challenge.

Train

You can use train.py for cycle-free CycleGAN.

For train.py, training input & target data, test input & target data folder directory is required.

Test

You can use inference.py for pre-trained cycle-free CycleGAN.

For inference.py, test input & target data folder directory and pre-trained weight file directory is required.